pax_global_header00006660000000000000000000000064147703371320014521gustar00rootroot0000000000000052 comment=f8079cd6ca6aa2f38dfc5490ad29333cea20309a beanquery-0.2.0/000077500000000000000000000000001477033713200135135ustar00rootroot00000000000000beanquery-0.2.0/.github/000077500000000000000000000000001477033713200150535ustar00rootroot00000000000000beanquery-0.2.0/.github/workflows/000077500000000000000000000000001477033713200171105ustar00rootroot00000000000000beanquery-0.2.0/.github/workflows/checks.yaml000066400000000000000000000032151477033713200212350ustar00rootroot00000000000000name: checks on: - push - pull_request jobs: ruff: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - uses: actions/setup-python@v5 with: python-version: '3.11' - run: python -m pip install ruff - run: ruff check beanquery/ coverage: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - uses: actions/setup-python@v5 with: python-version: '3.11' - run: pip install -r requirements.txt coverage - name: Check coverage by module # Check that each module is exaustively tested by the dedicated units tests. run: | set -x echo '{ "beanquery/parser/*": "beanquery/parser_test.py", "beanquery/query_render.py": "beanquery/query_render_test.py" }' | jq -rc 'to_entries | .[] | (.key + "=" + .value)' | while IFS='=' read src test do python -m coverage run --branch --include "$src" --omit beanquery/parser/parser.py -m unittest "$test" python -m coverage report --fail-under=100 -m python -m coverage erase done - name: Check overall coverage run: | python -m coverage run --branch -m unittest discover -t . -s beanquery/ -p \*_test.py python -m coverage report --sort cover build: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - uses: actions/setup-python@v5 with: python-version: '3.11' - run: pip install -r requirements.txt wheel build - run: python -m build --no-isolation - run: python -m pip install dist/beanquery-*.whl beanquery-0.2.0/.github/workflows/docs.yml000066400000000000000000000012111477033713200205560ustar00rootroot00000000000000name: docs on: push: pull_request: jobs: build: runs-on: ubuntu-latest steps: - uses: actions/setup-python@v5 with: python-version: '3.11' - uses: actions/checkout@v4 - run: python -m pip install .[docs] - run: python -m sphinx -W -b html docs/ build/html/ - uses: actions/upload-pages-artifact@v3 with: path: build/html deploy: needs: build permissions: pages: write id-token: write environment: name: github-pages runs-on: ubuntu-latest if: github.ref == 'refs/heads/master' steps: - uses: actions/deploy-pages@v4 beanquery-0.2.0/.github/workflows/release.yaml000066400000000000000000000011571477033713200214200ustar00rootroot00000000000000name: release on: push: tags: - 'v[0-9]*' jobs: build: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - run: python -m pip install build - run: python -m build - uses: actions/upload-artifact@v4 with: path: dist/* upload: needs: build runs-on: ubuntu-latest environment: upload permissions: id-token: write steps: - uses: actions/download-artifact@v4 with: merge-multiple: true path: dist - uses: pypa/gh-action-pypi-publish@release/v1 with: attestations: false beanquery-0.2.0/.github/workflows/test.yaml000066400000000000000000000016621477033713200207600ustar00rootroot00000000000000name: test on: - push - pull_request jobs: test: runs-on: ubuntu-latest strategy: fail-fast: false matrix: python: - '3.8' - '3.9' - '3.10' - '3.11' - '3.12' - '3.13' beancount: - '~= 2.3.6' - '~= 3.0.0' - '@ git+https://github.com/beancount/beancount.git' exclude: - python: '3.8' beancount: '@ git+https://github.com/beancount/beancount.git' - python: '3.9' beancount: '@ git+https://github.com/beancount/beancount.git' steps: - uses: actions/checkout@v4 - uses: actions/setup-python@v5 with: python-version: ${{ matrix.python }} allow-prereleases: true - run: pip install 'beancount ${{ matrix.beancount }}' - run: pip install -r requirements.txt - run: python -m unittest discover -p '*_test.py' beanquery-0.2.0/.gitignore000066400000000000000000000000151477033713200154770ustar00rootroot00000000000000__pycache__/ beanquery-0.2.0/CHANGES.rst000066400000000000000000000042541477033713200153220ustar00rootroot00000000000000Version 0.1 (unreleased) ------------------------ - The ``HAVING`` clause for aggregate queries is now supported. - The ``empty()`` BQL function to determine whether an Inventory object as returned by the ``sum()`` aggregate function is empty has been added. - Added the ``round()`` BQL function. - ``NULL`` values in ``SORT BY`` clause are now always considered to be smaller than any other values. This may results in rows to be returned in a slightly different order. - It is now possible to specify the direction of the ordering for each column in the ``SORT BY`` clause. This brings BQL closer to SQL specification but queries written with the old behaviour in mind will return rows in a different order. The query:: SELECT date, narration ORDER BY date, narration DESC used to return rows in descending order by both ``date`` and ``narration`` while now it would order the rows ascending by ``date`` and descending by ``narration``. To recover the old behavior, the query should be written:: SELECT date, narration ORDER BY date DESC, narration DESC - Type casting functions ``int()``, ``decimal()``, ``str()``, ``date()`` have been added. These are mostly useful to convert the generic ``object`` type returned by the metadata retrieval functions but can also be used to convert between types. If the conversion fails, ``NULL`` is returned. - The ``str()`` BQL function used to return a string representation of its argument using the Python :py:func:`repr()` function. This clashes with the use of ``str()`` as a type casting function. The function is renamed ``repr()``. - The ``date()`` BQL function used to extract a date from string arguments with a very relaxed parser. This clashes with the use of ``date()`` as a type casting function. The function is renamed ``parse_date()``. Another form of ``parse_date()`` that accepts the date format as second argument has been added. - The ``getitem()`` BQL function return type has been changed from a string to a generic ``object`` to match the return type of function retrieving entries from metadata dictionaries. The old behavior can be obtained with ``str(getitem(x, key))``. beanquery-0.2.0/LICENSE000066400000000000000000000432541477033713200145300ustar00rootroot00000000000000 GNU GENERAL PUBLIC LICENSE Version 2, June 1991 Copyright (C) 1989, 1991 Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. Preamble The licenses for most software are designed to take away your freedom to share and change it. By contrast, the GNU General Public License is intended to guarantee your freedom to share and change free software--to make sure the software is free for all its users. This General Public License applies to most of the Free Software Foundation's software and to any other program whose authors commit to using it. (Some other Free Software Foundation software is covered by the GNU Lesser General Public License instead.) You can apply it to your programs, too. When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things. To protect your rights, we need to make restrictions that forbid anyone to deny you these rights or to ask you to surrender the rights. These restrictions translate to certain responsibilities for you if you distribute copies of the software, or if you modify it. For example, if you distribute copies of such a program, whether gratis or for a fee, you must give the recipients all the rights that you have. You must make sure that they, too, receive or can get the source code. And you must show them these terms so they know their rights. We protect your rights with two steps: (1) copyright the software, and (2) offer you this license which gives you legal permission to copy, distribute and/or modify the software. Also, for each author's protection and ours, we want to make certain that everyone understands that there is no warranty for this free software. If the software is modified by someone else and passed on, we want its recipients to know that what they have is not the original, so that any problems introduced by others will not reflect on the original authors' reputations. Finally, any free program is threatened constantly by software patents. We wish to avoid the danger that redistributors of a free program will individually obtain patent licenses, in effect making the program proprietary. To prevent this, we have made it clear that any patent must be licensed for everyone's free use or not licensed at all. The precise terms and conditions for copying, distribution and modification follow. GNU GENERAL PUBLIC LICENSE TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 0. This License applies to any program or other work which contains a notice placed by the copyright holder saying it may be distributed under the terms of this General Public License. The "Program", below, refers to any such program or work, and a "work based on the Program" means either the Program or any derivative work under copyright law: that is to say, a work containing the Program or a portion of it, either verbatim or with modifications and/or translated into another language. (Hereinafter, translation is included without limitation in the term "modification".) Each licensee is addressed as "you". Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted, and the output from the Program is covered only if its contents constitute a work based on the Program (independent of having been made by running the Program). Whether that is true depends on what the Program does. 1. You may copy and distribute verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program. You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee. 2. You may modify your copy or copies of the Program or any portion of it, thus forming a work based on the Program, and copy and distribute such modifications or work under the terms of Section 1 above, provided that you also meet all of these conditions: a) You must cause the modified files to carry prominent notices stating that you changed the files and the date of any change. b) You must cause any work that you distribute or publish, that in whole or in part contains or is derived from the Program or any part thereof, to be licensed as a whole at no charge to all third parties under the terms of this License. c) If the modified program normally reads commands interactively when run, you must cause it, when started running for such interactive use in the most ordinary way, to print or display an announcement including an appropriate copyright notice and a notice that there is no warranty (or else, saying that you provide a warranty) and that users may redistribute the program under these conditions, and telling the user how to view a copy of this License. (Exception: if the Program itself is interactive but does not normally print such an announcement, your work based on the Program is not required to print an announcement.) These requirements apply to the modified work as a whole. If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it. Thus, it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program. In addition, mere aggregation of another work not based on the Program with the Program (or with a work based on the Program) on a volume of a storage or distribution medium does not bring the other work under the scope of this License. 3. You may copy and distribute the Program (or a work based on it, under Section 2) in object code or executable form under the terms of Sections 1 and 2 above provided that you also do one of the following: a) Accompany it with the complete corresponding machine-readable source code, which must be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, b) Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or, c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.) The source code for a work means the preferred form of the work for making modifications to it. For an executable work, complete source code means all the source code for all modules it contains, plus any associated interface definition files, plus the scripts used to control compilation and installation of the executable. However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable. If distribution of executable or object code is made by offering access to copy from a designated place, then offering equivalent access to copy the source code from the same place counts as distribution of the source code, even though third parties are not compelled to copy the source along with the object code. 4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. Any attempt otherwise to copy, modify, sublicense or distribute the Program is void, and will automatically terminate your rights under this License. However, parties who have received copies, or rights, from you under this License will not have their licenses terminated so long as such parties remain in full compliance. 5. You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it. 6. Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions. You may not impose any further restrictions on the recipients' exercise of the rights granted herein. You are not responsible for enforcing compliance by third parties to this License. 7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program. If any portion of this section is held invalid or unenforceable under any particular circumstance, the balance of the section is intended to apply and the section as a whole is intended to apply in other circumstances. It is not the purpose of this section to induce you to infringe any patents or other property right claims or to contest validity of any such claims; this section has the sole purpose of protecting the integrity of the free software distribution system, which is implemented by public license practices. Many people have made generous contributions to the wide range of software distributed through that system in reliance on consistent application of that system; it is up to the author/donor to decide if he or she is willing to distribute software through any other system and a licensee cannot impose that choice. This section is intended to make thoroughly clear what is believed to be a consequence of the rest of this License. 8. If the distribution and/or use of the Program is restricted in certain countries either by patents or by copyrighted interfaces, the original copyright holder who places the Program under this License may add an explicit geographical distribution limitation excluding those countries, so that distribution is permitted only in or among countries not thus excluded. In such case, this License incorporates the limitation as if written in the body of this License. 9. The Free Software Foundation may publish revised and/or new versions of the General Public License from time to time. Such new versions will be similar in spirit to the present version, but may differ in detail to address new problems or concerns. Each version is given a distinguishing version number. If the Program specifies a version number of this License which applies to it and "any later version", you have the option of following the terms and conditions either of that version or of any later version published by the Free Software Foundation. If the Program does not specify a version number of this License, you may choose any version ever published by the Free Software Foundation. 10. If you wish to incorporate parts of the Program into other free programs whose distribution conditions are different, write to the author to ask for permission. For software which is copyrighted by the Free Software Foundation, write to the Free Software Foundation; we sometimes make exceptions for this. Our decision will be guided by the two goals of preserving the free status of all derivatives of our free software and of promoting the sharing and reuse of software generally. NO WARRANTY 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. END OF TERMS AND CONDITIONS How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it free software which everyone can redistribute and change under these terms. To do so, attach the following notices to the program. It is safest to attach them to the start of each source file to most effectively convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. Copyright (C) This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: Gnomovision version 69, Copyright (C) year name of author Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. The hypothetical commands `show w' and `show c' should show the appropriate parts of the General Public License. Of course, the commands you use may be called something other than `show w' and `show c'; they could even be mouse-clicks or menu items--whatever suits your program. You should also get your employer (if you work as a programmer) or your school, if any, to sign a "copyright disclaimer" for the program, if necessary. Here is a sample; alter the names: Yoyodyne, Inc., hereby disclaims all copyright interest in the program `Gnomovision' (which makes passes at compilers) written by James Hacker. , 1 April 1989 Ty Coon, President of Vice This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Lesser General Public License instead of this License. beanquery-0.2.0/README.rst000066400000000000000000000004171477033713200152040ustar00rootroot00000000000000beanquery: Customizable lightweight SQL query tool ================================================== beanquery is a customizable and extensible lightweight SQL query tool that works on tabular data, including `Beancount`__ ledger data. __ https://beancount.github.io/ beanquery-0.2.0/beanquery/000077500000000000000000000000001477033713200155065ustar00rootroot00000000000000beanquery-0.2.0/beanquery/__init__.py000066400000000000000000000042041477033713200176170ustar00rootroot00000000000000import importlib from urllib.parse import urlparse from . import parser from . import compiler from . import tables from .compiler import CompilationError from .cursor import Cursor, Column from .errors import Warning, Error, InterfaceError, DatabaseError, DataError, OperationalError from .errors import IntegrityError, InternalError, ProgrammingError, NotSupportedError from .parser import ParseError __version__ = '0.2.0' # DB-API compliance apilevel = '2.0' threadsafety = 2 paramstyle = 'pyformat' def connect(dsn, **kwargs): return Connection(dsn, **kwargs) class Connection: def __init__(self, dsn='', **kwargs): self.tables = {None: tables.NullTable()} # The ``None`` table is the default table. The ``''`` table is the # table that is explicitly selected with ``FROM #``. Having the # default table and the ``''`` table allows to select the empty table # when the ``beancount`` data source is initialized and it sets the # default table to the ``postings`` table. self.tables[''] = self.tables[None] # These are used only by the ``beancount`` data source. self.options = {} self.errors = [] if dsn: self.attach(dsn, **kwargs) def attach(self, dsn, **kwargs): scheme = urlparse(dsn).scheme source = importlib.import_module(f'beanquery.sources.{scheme}') source.attach(self, dsn, **kwargs) def close(self): # Required by the DB-API. pass def parse(self, query): return parser.parse(query) def compile(self, query): return compiler.compile(self, query) def execute(self, query, params=None): return self.cursor().execute(query, params) def cursor(self): return Cursor(self) __all__ = [ 'Column', 'CompilationError', 'Connection', 'Cursor', 'DataError', 'DatabaseError', 'Error', 'IntegrityError', 'InterfaceError', 'InternalError', 'NotSupportedError', 'OperationalError', 'ParseError', 'ProgrammingError', 'Warning', 'apilevel', 'connet', 'paramstyle', 'threadsafety', ] beanquery-0.2.0/beanquery/__main__.py000066400000000000000000000001111477033713200175710ustar00rootroot00000000000000from beanquery import shell if __name__ == '__main__': shell.main() beanquery-0.2.0/beanquery/compiler.py000066400000000000000000001106361477033713200177010ustar00rootroot00000000000000import collections.abc import importlib import typing from decimal import Decimal from functools import singledispatchmethod from os import path from typing import Optional, Sequence, Mapping, Union from urllib.parse import urlparse from . import types from . import parser from .errors import ProgrammingError from .parser import ast from .query_compile import ( EvalAggregator, EvalAnd, EvalAll, EvalAny, EvalCoalesce, EvalColumn, EvalConstant, EvalCreateTable, EvalGetItem, EvalGetter, EvalInsert, EvalOr, EvalPivot, EvalQuery, EvalConstantSubquery1D, EvalRow, EvalTarget, FUNCTIONS, OPERATORS, SubqueryTable, ) # Load functions and types definitions. from . import query_env # noqa: F401 # A global constant which sets whether we support inferred/implicit group-by # semantics. SUPPORT_IMPLICIT_GROUPBY = True class CompilationError(ProgrammingError): def __init__(self, message, node=None): super().__init__(message) self.parseinfo = node.parseinfo if node is not None else None class Compiler: def __init__(self, context): self.context = context self.stack = [context.tables.get(None)] @property def table(self): return self.stack[-1] @table.setter def table(self, value): self.stack[-1] = value def compile(self, query, parameters=None): """Compile an AST into an executable statement.""" self.parameters = parameters placeholders = [node for node in query.walk() if isinstance(node, ast.Placeholder)] if placeholders: names = {placeholder.name for placeholder in placeholders} if all(names): if not isinstance(parameters, Mapping): raise TypeError('query parameters should be a mapping when using named placeholders') if names - parameters.keys(): missing = ', '.join(sorted(names - parameters.keys())) raise ProgrammingError(f'query parameter missing: {missing}') elif not any(names): if not isinstance(parameters, Sequence): raise TypeError('query parameters should be a sequence when using positional placeholders') if len(placeholders) != len(parameters): raise ProgrammingError( f'the query has {len(placeholders)} placeholders but {len(parameters)} parameters were passed') for i, placeholder in enumerate(sorted(placeholders, key=lambda node: node.parseinfo.pos)): placeholder.name = i else: raise ProgrammingError('positional and named parameters cannot be mixed') return self._compile(query) @singledispatchmethod def _compile(self, node: Optional[ast.Node]): if node is None: return None raise NotImplementedError @_compile.register def _select(self, node: ast.Select): self.stack.append(self.table) # Compile the FROM clause. c_from_expr = self._compile_from(node.from_clause) # Compile the targets. c_targets = self._compile_targets(node.targets) # Bind the WHERE expression to the execution environment. c_where = self._compile(node.where_clause) # Check that the FROM clause does not contain aggregates. This # should never trigger if the compilation environment does not # contain any aggregate. if c_where is not None and is_aggregate(c_where): raise CompilationError('aggregates are not allowed in WHERE clause') # Combine FROM and WHERE clauses if c_from_expr is not None: c_where = c_from_expr if c_where is None else EvalAnd([c_from_expr, c_where]) # Process the GROUP-BY clause. new_targets, group_indexes, having_index = self._compile_group_by(node.group_by, c_targets) c_targets.extend(new_targets) # Process the ORDER-BY clause. new_targets, order_spec = self._compile_order_by(node.order_by, c_targets) c_targets.extend(new_targets) # If this is an aggregate query (it groups, see list of indexes), check that # the set of non-aggregates match exactly the group indexes. This should # always be the case at this point, because we have added all the necessary # targets to the list of group-by expressions and should have resolved all # the indexes. if group_indexes is not None: non_aggregate_indexes = {index for index, c_target in enumerate(c_targets) if not c_target.is_aggregate} if non_aggregate_indexes != set(group_indexes): missing_names = ['"{}"'.format(c_targets[index].name) for index in non_aggregate_indexes - set(group_indexes)] raise CompilationError( 'all non-aggregates must be covered by GROUP-BY clause in aggregate query: ' 'the following targets are missing: {}'.format(','.join(missing_names))) query = EvalQuery(self.table, c_targets, c_where, group_indexes, having_index, order_spec, node.limit, node.distinct) pivots = self._compile_pivot_by(node.pivot_by, c_targets, group_indexes) if pivots: return EvalPivot(query, pivots) self.stack.pop() return query def _compile_from(self, node): if node is None: return None # Subquery. if isinstance(node, ast.Select): self.table = SubqueryTable(self._compile(node)) return None # Table reference. if isinstance(node, ast.Table): self.table = self.context.tables.get(node.name) if self.table is None: raise CompilationError(f'table "{node.name}" does not exist', node) return None # FROM expression. if isinstance(node, ast.From): # Check if the FROM expression is a column name belongin to the current table. if isinstance(node.expression, ast.Column): column = self.table.columns.get(node.expression.name) if column is None: # When it is not, threat it as a table name. table = self.context.tables.get(node.expression.name) if table is not None: self.table = table return None c_expression = self._compile(node.expression) # Check that the FROM clause does not contain aggregates. if c_expression is not None and is_aggregate(c_expression): raise CompilationError('aggregates are not allowed in FROM clause') if node.open and node.close and node.open > node.close: raise CompilationError('CLOSE date must follow OPEN date') # Apply OPEN, CLOSE, and CLEAR clauses. if node.open is not None or node.close is not None or node.clear is not None: self.table = self.table.evolve(open=node.open, close=node.close, clear=node.clear) return c_expression raise NotImplementedError def _compile_targets(self, targets): """Compile the targets and check for their validity. Process wildcard. Args: targets: A list of target expressions from the parser. Returns: A list of compiled target expressions with resolved names. """ # Bind the targets expressions to the execution context. if isinstance(targets, ast.Asterisk): # Insert the full list of available columns. targets = [ast.Target(ast.Column(name), None) for name in self.table.wildcard_columns] # Compile targets. c_targets = [] for target in targets: c_expr = self._compile(target.expression) name = get_target_name(target) c_targets.append(EvalTarget(c_expr, name, is_aggregate(c_expr))) columns, aggregates = get_columns_and_aggregates(c_expr) # Check for mixed aggregates and non-aggregates. if columns and aggregates: raise CompilationError('mixed aggregates and non-aggregates are not allowed') # Check for aggregates of aggregates. for aggregate in aggregates: for child in aggregate.childnodes(): if is_aggregate(child): raise CompilationError('aggregates of aggregates are not allowed') return c_targets def _compile_order_by(self, order_by, c_targets): """Process an order-by clause. Args: order_by: A OrderBy instance as provided by the parser. c_targets: A list of compiled target expressions. Returns: A tuple of new_targets: A list of new compiled target nodes. order_spec: A list of (integer indexes, sort order) tuples. """ if not order_by: return [], None # Compile order-by expressions and resolve them to their targets if # possible. A ORDER-BY column may be one of the following: # # * A reference to a target by name. # * A reference to a target by index (starting at one). # * A new expression, aggregate or not. # # References by name are converted to indexes. New expressions are # inserted into the list of targets as invisible targets. new_targets = c_targets[:] c_target_expressions = [c_target.c_expr for c_target in c_targets] targets_name_map = {target.name: idx for idx, target in enumerate(c_targets) if target.name is not None} # Only targets appearing in the SELECT targets list can be # referenced by index. These are guaranteed to have a valid name. n_targets = len(targets_name_map) order_spec = [] for spec in order_by: column = spec.column descending = spec.ordering index = None # Process target references by index. if isinstance(column, int): index = column - 1 if not 0 <= index < n_targets: raise CompilationError(f'invalid ORDER-BY column index {column}') else: # Process target references by name. These will be parsed as # simple Column expressions. If they refer to a target name, we # resolve them. if isinstance(column, ast.Column): name = column.name index = targets_name_map.get(name, None) # Otherwise we compile the expression and add it to the list of # targets to evaluate and index into that new target. if index is None: c_expr = self._compile(column) # Attempt to reconcile the expression with one of the existing # target expressions. try: index = c_target_expressions.index(c_expr) except ValueError: # Add the new target. 'None' for the target name implies it # should be invisible, not to be rendered. index = len(new_targets) new_targets.append(EvalTarget(c_expr, None, is_aggregate(c_expr))) c_target_expressions.append(c_expr) assert index is not None, "Internal error, could not index order-by reference." order_spec.append((index, descending)) return new_targets[len(c_targets):], order_spec def _compile_pivot_by(self, pivot_by, targets, group_indexes): """Compiles a PIVOT BY clause. Resolve and validate columns references in the PIVOT BY clause. The PIVOT BY clause accepts two name od index references to columns in the SELECT targets list. The second columns should be a GROUP BY column so that the values of the pivot column are unique. """ if pivot_by is None: return None indexes = [] names = {target.name: index for index, target in enumerate(targets)} for column in pivot_by.columns: # Process target references by index. if isinstance(column, int): index = column - 1 if not 0 <= index < len(targets): raise CompilationError(f'invalid PIVOT BY column index {column}') indexes.append(index) continue # Process target references by name. if isinstance(column, ast.Column): index = names.get(column.name, None) if index is None: raise CompilationError(f'PIVOT BY column {column!r} is not in the targets list') indexes.append(index) continue # Not reached. raise RuntimeError # Sanity checks. if indexes[0] == indexes[1]: raise CompilationError('the two PIVOT BY columns cannot be the same column') if indexes[1] not in group_indexes: raise CompilationError('the second PIVOT BY column must be a GROUP BY column') return indexes def _compile_group_by(self, group_by, c_targets): """Process a group-by clause. Args: group_by: A GroupBy instance as provided by the parser. c_targets: A list of compiled target expressions. Returns: A tuple of new_targets: A list of new compiled target nodes. group_indexes: If the query is an aggregate query, a list of integer indexes to be used for processing grouping. Note that this list may be empty (in the case of targets with only aggregates). On the other hand, if this is not an aggregated query, this is set to None. So do distinguish the empty list vs. None. """ new_targets = c_targets[:] c_target_expressions = [c_target.c_expr for c_target in c_targets] group_indexes = [] having_index = None if group_by: assert group_by.columns, "Internal error with GROUP-BY parsing" # Compile group-by expressions and resolve them to their targets if # possible. A GROUP-BY column may be one of the following: # # * A reference to a target by name. # * A reference to a target by index (starting at one). # * A new, non-aggregate expression. # # References by name are converted to indexes. New expressions are # inserted into the list of targets as invisible targets. targets_name_map = {target.name: index for index, target in enumerate(c_targets)} for column in group_by.columns: index = None # Process target references by index. if isinstance(column, int): index = column - 1 if not 0 <= index < len(c_targets): raise CompilationError(f'invalid GROUP-BY column index {column}') else: # Process target references by name. These will be parsed as # simple Column expressions. If they refer to a target name, we # resolve them. if isinstance(column, ast.Column): name = column.name index = targets_name_map.get(name, None) # Otherwise we compile the expression and add it to the list of # targets to evaluate and index into that new target. if index is None: c_expr = self._compile(column) # Check if the new expression is an aggregate. aggregate = is_aggregate(c_expr) if aggregate: raise CompilationError(f'GROUP-BY expressions may not be aggregates: "{column}"') # Attempt to reconcile the expression with one of the existing # target expressions. try: index = c_target_expressions.index(c_expr) except ValueError: # Add the new target. 'None' for the target name implies it # should be invisible, not to be rendered. index = len(new_targets) new_targets.append(EvalTarget(c_expr, None, aggregate)) c_target_expressions.append(c_expr) assert index is not None, "Internal error, could not index group-by reference." group_indexes.append(index) # Check that the group-by column references a non-aggregate. c_expr = new_targets[index].c_expr if is_aggregate(c_expr): raise CompilationError(f'GROUP-BY expressions may not reference aggregates: "{column}"') # Check that the group-by column has a supported hashable type. if not issubclass(c_expr.dtype, collections.abc.Hashable): raise CompilationError(f'GROUP-BY a non-hashable type is not supported: "{column}"') # Compile HAVING clause. if group_by.having is not None: c_expr = self._compile(group_by.having) if not is_aggregate(c_expr): raise CompilationError('the HAVING clause must be an aggregate expression') having_index = len(new_targets) new_targets.append(EvalTarget(c_expr, None, True)) c_target_expressions.append(c_expr) else: # If it does not have a GROUP-BY clause... aggregate_bools = [c_target.is_aggregate for c_target in c_targets] if any(aggregate_bools): # If the query is an aggregate query, check that all the targets are # aggregates. if all(aggregate_bools): # FIXME: shold we really be checking for the empty # list or is checking for a false value enough? assert group_indexes == [] elif SUPPORT_IMPLICIT_GROUPBY: # If some of the targets aren't aggregates, automatically infer # that they are to be implicit group by targets. This makes for # a much more convenient syntax for our lightweight SQL, where # grouping is optional. group_indexes = [ index for index, c_target in enumerate(c_targets) if not c_target.is_aggregate] else: raise CompilationError('aggregate query without a GROUP-BY should have only aggregates') else: # This is not an aggregate query; don't set group_indexes to # anything useful, we won't need it. group_indexes = None return new_targets[len(c_targets):], group_indexes, having_index @_compile.register def _column(self, node: ast.Column): column = self.table.columns.get(node.name) if column is not None: return column raise CompilationError(f'column "{node.name}" not found in table "{self.table.name}"', node) @_compile.register def _or(self, node: ast.Or): return EvalOr([self._compile(arg) for arg in node.args]) @_compile.register def _and(self, node: ast.And): return EvalAnd([self._compile(arg) for arg in node.args]) _OPERATORS = { '<': ast.Less, '<=': ast.LessEq, '>': ast.Greater, '>=': ast.GreaterEq, '=': ast.Equal, '!=': ast.NotEqual, '~': ast.Match, '!~': ast.NotMatch, '?~': ast.Matches, } # dispatching on an Union is supported only starting with Python 3.11 @_compile.register(ast.All) @_compile.register(ast.Any) def _all(self, node): right = self._compile(node.right) if isinstance(right, EvalQuery): if len(right.columns) != 1: raise CompilationError('subquery has too many columns', node.right) right = EvalConstantSubquery1D(right) right_dtype = typing.get_origin(right.dtype) or right.dtype if right_dtype not in {list, set}: raise CompilationError(f'not a list or set but {right_dtype}', node.right) args = typing.get_args(right.dtype) if args: assert len(args) == 1 right_element_dtype = args[0] else: right_element_dtype = object left = self._compile(node.left) # lookup operator implementaton and check typing op = self._OPERATORS[node.op] for func in OPERATORS[op]: if func.__intypes__ == [right_element_dtype, left.dtype]: break else: raise CompilationError( f'operator "{op.__name__.lower()}(' f'{left.dtype.__name__}, {right_element_dtype.__name__})" not supported', node) # need to instantiate the operaotr implementation to get to the underlying function operator = func(None, None).operator cls = EvalAll if type(node) is ast.All else EvalAny return cls(operator, left, right) @_compile.register def _function(self, node: ast.Function): operands = [self._compile(operand) for operand in node.operands] # ``row(*)`` is parsed like a function call but does something special if node.fname == 'row' and len(operands) == 1 and operands[0].dtype == types.Asterisk: return EvalRow() # ``coalesce()`` is parsed like a function call but it does # not really fit our model for function evaluation, therefore # it gets special threatment here. if node.fname == 'coalesce': for operand in operands: if operand.dtype != operands[0].dtype: dtypes = ', '.join(operand.dtype.__name__ for operand in operands) raise CompilationError(f'coalesce() function arguments must have uniform type, found: {dtypes}', node) return EvalCoalesce(operands) function = types.function_lookup(FUNCTIONS, node.fname, operands) if function is None: sig = '{}({})'.format(node.fname, ', '.join(f'{operand.dtype.__name__.lower()}' for operand in operands)) raise CompilationError(f'no function matches "{sig}" name and argument types', node) # Replace ``meta(key)`` with ``meta[key]``. if node.fname == 'meta': key = node.operands[0] node = ast.Function('getitem', [ast.Column('meta', parseinfo=node.parseinfo), key]) return self._compile(node) # Replace ``entry_meta(key)`` with ``entry.meta[key]``. if node.fname == 'entry_meta': key = node.operands[0] node = ast.Function('getitem', [ast.Attribute(ast.Column('entry', parseinfo=node.parseinfo), 'meta'), key]) return self._compile(node) # Replace ``any_meta(key)`` with ``getitem(meta, key, entry.meta[key])``. if node.fname == 'any_meta': key = node.operands[0] node = ast.Function('getitem', [ast.Column('meta', parseinfo=node.parseinfo), key, ast.Function('getitem', [ ast.Attribute(ast.Column('entry', parseinfo=node.parseinfo), 'meta'), key])]) return self._compile(node) # Replace ``has_account(regexp)`` with ``('(?i)' + regexp) ~? any (accounts)``. if node.fname == 'has_account': node = ast.Any(ast.Add(ast.Constant('(?i)'), node.operands[0]), '?~', ast.Column('accounts')) return self._compile(node) function = function(self.context, operands) # Constants folding. if all(isinstance(operand, EvalConstant) for operand in operands) and function.pure: return EvalConstant(function(None), function.dtype) return function @_compile.register def _subscript(self, node: ast.Subscript): operand = self._compile(node.operand) if issubclass(operand.dtype, dict): return EvalGetItem(operand, node.key) raise CompilationError('column type is not subscriptable', node) @_compile.register def _attribute(self, node: ast.Attribute): operand = self._compile(node.operand) dtype = types.ALIASES.get(operand.dtype, operand.dtype) if issubclass(dtype, types.Structure): getter = dtype.columns.get(node.name) if getter is None: raise CompilationError(f'structured type has no attribute "{node.name}"', node) return EvalGetter(operand, getter, getter.dtype) raise CompilationError('column type is not structured', node) @_compile.register def _unaryop(self, node: ast.UnaryOp): operand = self._compile(node.operand) function = types.function_lookup(OPERATORS, type(node), [operand]) if function is None: raise CompilationError( f'operator "{type(node).__name__.lower()}({types.name(operand.dtype)})" not supported', node) function = function(operand) # Constants folding. if isinstance(operand, EvalConstant): return EvalConstant(function(None), function.dtype) return function @_compile.register def _between(self, node: ast.Between): operand = self._compile(node.operand) lower = self._compile(node.lower) upper = self._compile(node.upper) intypes = [operand.dtype, lower.dtype, upper.dtype] for candidate in OPERATORS[type(node)]: if candidate.__intypes__ == intypes: func = candidate(operand, lower, upper) return func raise CompilationError( f'operator "{types.name(operand.dtype)} BETWEEN {types.name(lower.dtype)} ' f'AND {types.name(upper.dtype)}" not supported', node) @_compile.register(ast.In) @_compile.register(ast.NotIn) def _inop(self, node: Union[ast.In, ast.NotIn]): left = self._compile(node.left) right = self._compile(node.right) if isinstance(right, EvalQuery): if len(right.columns) != 1: raise CompilationError('subquery has too many columns', node.right) right = EvalConstantSubquery1D(right) op = OPERATORS[type(node)][0] return op(left, right) @_compile.register def _binaryop(self, node: ast.BinaryOp): left = self._compile(node.left) right = self._compile(node.right) candidates = OPERATORS[type(node)] while True: intypes = [left.dtype, right.dtype] for op in candidates: if op.__intypes__ == intypes: function = op(left, right) # Constants folding. if isinstance(left, EvalConstant) and isinstance(right, EvalConstant): return EvalConstant(function(None), function.dtype) return function # Implement type inference when one of the operands is not strongly typed. if left.dtype is object and right.dtype is not object: target = right.dtype if target is int: # The Beancount parser does not emit int typed # values, thus casting to int is only going to # loose information. Promote to decimal. target = Decimal name = types.MAP.get(target) if name is None: break left = types.function_lookup(FUNCTIONS, name, [left])(self.context, [left]) continue if right.dtype is object and left.dtype is not object: target = left.dtype if target is int: # The Beancount parser does not emit int typed # values, thus casting to int is only going to # loose information. Promote to decimal. target = Decimal name = types.MAP.get(target) if name is None: break right = types.function_lookup(FUNCTIONS, name, [right])(self.context, [right]) continue # Failure. break raise CompilationError( f'operator "{type(node).__name__.lower()}(' f'{types.name(left.dtype)}, {types.name(right.dtype)})" not supported', node) @_compile.register def _constant(self, node: ast.Constant): # For backward compatibility, the parser allows strings to be # delimited by single or double quotes. This creates ambiguity between # quoted identifiers and string. Strings delimited by double quotes # are treated as column names when they resolve to an existing column # in the current table. if isinstance(node.value, str) and node.text and node.text[0] == '"': if node.value in self.table.columns: return self._column(ast.Column(node.value)) return EvalConstant(node.value) @_compile.register def _placeholder(self, node: ast.Placeholder): return EvalConstant(self.parameters[node.name]) @_compile.register def _asterisk(self, node: ast.Asterisk): return EvalConstant(None, dtype=types.Asterisk) @_compile.register def _balances(self, node: ast.Balances): return self._compile(transform_balances(node)) @_compile.register def _journal(self, node: ast.Journal): return self._compile(transform_journal(node)) @_compile.register def _print(self, node: ast.Print): self.table = self.context.tables.get('entries') expr = self._compile_from(node.from_clause) targets = [EvalTarget(EvalRow(), 'ROW(*)', False)] return EvalQuery(self.table, targets, expr, None, None, None, None, False) @_compile.register def _create_table(self, node: ast.CreateTable): query = None columns = None if node.columns is not None: columns = [] for cname, ctype in node.columns: datatype = types.parse(ctype) if datatype is None: raise CompilationError(f'unrecognized type "{ctype}"', node) columns.append((cname, datatype)) if node.query is not None: query = self._compile(node.query) columns = [(t.name, t.c_expr.dtype) for t in query.c_targets if t.name is not None] parts = urlparse(node.using) scheme = parts.scheme or path.splitext(parts.path)[1][1:] or 'memory' impl = importlib.import_module(f'beanquery.sources.{scheme}').create return EvalCreateTable(self.context, node.name, columns, node.using, query, impl) @_compile.register def _insert(self, node: ast.Insert): table = self.context.tables.get(node.table.name) if table is None: raise CompilationError(f'table "{node.table.name}" does not exist', node.table) impl = getattr(table, 'insert', None) if impl is None: raise CompilationError(f'table "{node.table.name}" does not support insertion', node.table) if len(node.values) != len(node.columns): raise CompilationError( f'column names and values mismatch: ' f'expected {len(node.columns)} but {len(node.values)} values were supplied', node) values = [EvalConstant(None)] * len(table.columns) columns = {name: i for i, name in enumerate(table.columns.keys())} for column, value in zip(node.columns, node.values): index = columns.get(column.name) if index is None: raise CompilationError(f'column "{column.name}" not found in table "{node.table.name}"', column) expr = self._compile(value) if not expr.dtype == table.columns.get(column.name).dtype: raise CompilationError(f'expression has wrong type for column "{column.name}"', value) values[index] = expr return EvalInsert(table, values) def transform_journal(journal): """Translate a Journal entry into an uncompiled Select statement. Args: journal: An instance of a Journal object. Returns: An instance of an uncompiled Select object. """ cooked_select = parser.parse(""" SELECT date, flag, MAXWIDTH(payee, 48), MAXWIDTH(narration, 80), account, {summary_func}(position), {summary_func}(balance) {where} """.format(where=('WHERE account ~ "{}"'.format(journal.account) if journal.account else ''), summary_func=journal.summary_func or '')) return ast.Select(cooked_select.targets, journal.from_clause, cooked_select.where_clause, None, None, None, None, None) def transform_balances(balances): """Translate a Balances entry into an uncompiled Select statement. Args: balances: An instance of a Balance object. Returns: An instance of an uncompiled Select object. """ ## FIXME: Change the aggregation rules to allow GROUP-BY not to include the ## non-aggregate ORDER-BY columns, so we could just GROUP-BY accounts here ## instead of having to include the sort-key. I think it should be fine if ## the first or last sort-order value gets used, because it would simplify ## the input statement. cooked_select = parser.parse(""" SELECT account, SUM({}(position)) GROUP BY account, ACCOUNT_SORTKEY(account) ORDER BY ACCOUNT_SORTKEY(account) """.format(balances.summary_func or "")) return ast.Select(cooked_select.targets, balances.from_clause, balances.where_clause, cooked_select.group_by, cooked_select.order_by, None, None, None) def get_target_name(target): """Compute the target name. This uses the same algorithm used by SQLite. If the target has an AS clause assigning it a name, that will be the name used. If the target refers directly to a column, then the target name is the column name. Otherwise use the expression text. """ if target.name is not None: return target.name if isinstance(target.expression, ast.Column): return target.expression.name return target.expression.text.strip() def get_columns_and_aggregates(node): """Find the columns and aggregate nodes below this tree. All nodes under aggregate nodes are ignored. Args: node: An instance of EvalNode. Returns: A pair of (columns, aggregates), both of which are lists of EvalNode instances. columns: The list of all columns accessed not under an aggregate node. aggregates: The list of all aggregate nodes. """ columns = [] aggregates = [] _get_columns_and_aggregates(node, columns, aggregates) return columns, aggregates def _get_columns_and_aggregates(node, columns, aggregates): """Walk down a tree of nodes and fetch the column accessors and aggregates. This function ignores all nodes under aggregate nodes. Args: node: An instance of EvalNode. columns: An accumulator for columns found so far. aggregate: An accumulator for aggregate notes found so far. """ if isinstance(node, EvalAggregator): aggregates.append(node) elif isinstance(node, EvalColumn): columns.append(node) else: for child in node.childnodes(): _get_columns_and_aggregates(child, columns, aggregates) def is_aggregate(node): """Return true if the node is an aggregate. Args: node: An instance of EvalNode. Returns: A boolean. """ # Note: We could be a tiny bit more efficient here, but it doesn't matter # much. Performance of the query compilation matters very little overall. _, aggregates = get_columns_and_aggregates(node) return bool(aggregates) def compile(context, statement, parameters=None): return Compiler(context).compile(statement, parameters) beanquery-0.2.0/beanquery/cursor.py000066400000000000000000000075461477033713200174110ustar00rootroot00000000000000from operator import attrgetter from typing import Sequence from . import types from . import parser from . import compiler class Column(Sequence): __module__ = 'beanquery' def __init__(self, name, datatype): self._name = name self._type = datatype _vars = tuple(attrgetter(name) for name in 'name type_code display_size internal_size precision scale null_ok'.split()) def __eq__(self, other): if isinstance(other, type(self)): return tuple(self) == tuple(other) if isinstance(other, tuple): # Used in tests. return (self._name, self._type) == other return NotImplemented def __repr__(self): return f'{self.__module__}.{self.__class__.__name__}({self._name!r}, {types.name(self._type)})' def __len__(self): return 7 def __getitem__(self, key): if isinstance(key, slice): return tuple(getter(self) for getter in self._vars(key)) return self._vars[key](self) @property def name(self): return self._name @property def datatype(self): # Extension to the DB-API. return self._type @property def type_code(self): # The DB-API specification is vague on this point, but other # database connection libraries expose this as an int. It does # not make much sense to keep a mapping between int type code # and actual types, thus just return the hash of the type # object. return hash(self._type) @property def display_size(self): return None @property def internal_size(self): return None @property def precision(self): return None @property def scale(self): return None @property def null_ok(self): return None class Cursor: def __init__(self, connection): self._context = connection self._description = None self._rows = None self._pos = 0 self.arraysize = 1 @property def connection(self): return self._context def execute(self, query, params=None): if not isinstance(query, parser.ast.Node): query = parser.parse(query) query = compiler.compile(self._context, query, params) description, rows = query() self._description = description self._rows = rows self._pos = 0 return self def executemany(self, query, params=None): query = parser.parse(query) for p in params: self.execute(query, p) @property def description(self): return self._description @property def rowcount(self): return len(self._rows) if self._rows is not None else -1 @property def rownumber(self): return self._pos def fetchone(self): # This implementation pops items from the front of the results # rows list and is thus not efficient, especially for large # results sets. if self._rows is None or not len(self._rows): return None self._pos += 1 return self._rows.pop(0) def fetchmany(self, size=None): if self._rows is None: return [] n = size if size is not None else self.arraysize rows = self._rows[:n] self._rows = self._rows[n:] self._pos += len(rows) return rows def fetchall(self): if self._rows is None: return [] rows = self._rows self._rows = [] self._pos += len(rows) return rows def close(self): # Required by the DB-API. pass def setinputsizes(self, sizes): # Required by the DB-API. pass def setoutputsize(self, size, column=None): # Required by the DB-API. pass def __iter__(self): return iter(self._rows if self._rows is not None else []) beanquery-0.2.0/beanquery/cursor_test.py000066400000000000000000000051731477033713200204420ustar00rootroot00000000000000import unittest import sqlite3 import beanquery from beanquery.sources import test class APITests: def test_description(self): curs = self.conn.cursor() self.assertIsNone(curs.description) curs.execute(f'SELECT x FROM {self.table} WHERE x = 0') self.assertEqual([c[0] for c in curs.description], ['x']) column = curs.description[0] self.assertEqual(len(column), 7) def test_cursor_not_initialized(self): curs = self.conn.cursor() self.assertIsNone(curs.fetchone()) self.assertEqual(curs.fetchmany(), []) self.assertEqual(curs.fetchall(), []) def test_cursor_fetchone(self): curs = self.conn.cursor() curs.execute(f'SELECT x FROM {self.table} WHERE x < 2') row = curs.fetchone() self.assertEqual(row, (0, )) row = curs.fetchone() self.assertEqual(row, (1, )) row = curs.fetchone() self.assertIsNone(row) def test_cursor_fetchall(self): curs = self.conn.cursor() curs.execute(f'SELECT x FROM {self.table} WHERE x < 2') rows = curs.fetchall() self.assertEqual(rows, [(0, ), (1, )]) rows = curs.fetchall() self.assertEqual(rows, []) def test_cursor_fethmany(self): curs = self.conn.cursor() curs.execute(f'SELECT x FROM {self.table} WHERE x < 2') rows = curs.fetchmany() self.assertEqual(rows, [(0, )]) rows = curs.fetchmany() self.assertEqual(rows, [(1, )]) rows = curs.fetchmany() self.assertEqual(rows, []) def test_cursor_iterator(self): curs = self.conn.cursor() o = object() row = next(iter(curs), o) self.assertIs(row, o) curs = self.conn.cursor() curs.execute(f'SELECT x FROM {self.table} WHERE x < 2') iterator = iter(curs) row = next(iterator) self.assertEqual(row, (0, )) row = next(iterator) self.assertEqual(row, (1, )) row = next(iterator, o) self.assertIs(row, o) class TestSQLite(APITests, unittest.TestCase): @classmethod def setUpClass(cls): cls.table = 'test' cls.conn = sqlite3.connect(':memory:') curs = cls.conn.cursor() curs.execute('CREATE TABLE test (x int)') curs.executemany('INSERT INTO test VALUES (?)', [(i, ) for i in range(16)]) @classmethod def tearDownClass(cls): cls.conn.close() class TestBeanquery(APITests, unittest.TestCase): @classmethod def setUpClass(cls): cls.table = '#test' cls.conn = beanquery.Connection() cls.conn.tables['test'] = test.Table(16) beanquery-0.2.0/beanquery/errors.py000066400000000000000000000031051477033713200173730ustar00rootroot00000000000000""" Exceptions hierarchy defined by the DB-API: Exception Warning Error InterfaceError DatabaseError DataError OperationalError IntegrityError InternalError ProgrammingError NotSupportedError """ class Warning(Exception): """Exception raised for important warnings.""" __module__ = 'beanquery' class Error(Exception): """Base exception for all errors.""" __module__ = 'beanquery' class InterfaceError(Error): """An error related to the database interface rather than the database itself.""" __module__ = 'beanquery' class DatabaseError(Error): """Exception raised for errors that are related to the database.""" __module__ = 'beanquery' class DataError(DatabaseError): """An error caused by problems with the processed data.""" __module__ = 'beanquery' class OperationalError(DatabaseError): """An error related to the database's operation.""" __module__ = 'beanquery' class IntegrityError(DatabaseError): """An error caused when the relational integrity of the database is affected.""" __module__ = 'beanquery' class InternalError(DatabaseError): """An error generated when the database encounters an internal error.""" __module__ = 'beanquery' class ProgrammingError(DatabaseError): """Exception raised for programming errors.""" __module__ = 'beanquery' class NotSupportedError(DatabaseError): """A method or database API was used which is not supported by the database.""" __module__ = 'beanquery' beanquery-0.2.0/beanquery/hashable.py000066400000000000000000000045021477033713200176300ustar00rootroot00000000000000import decimal import pickle import textwrap from . import types # Hashable types. Checking ``issubclass(T, types.Hashable)`` does not # work because named tuples pass that test and beancount uses many named # tuples that have dictionary members making them effectively not # hashable. FUNDAMENTAL = frozenset({ bool, bytes, complex, decimal.Decimal, float, int, str, # These are hashable only if the contained objects are hashable. frozenset, tuple, }) # Function reducing non-hashable types to something hashable. REDUCERS = {} def register(datatype, func): """Register reduce function for non-hashable type. The reduce function maps an non-hashable object into an hashable representation. This representation does not need to capture all the object facets, but it should retrurn someting unique enough to avoid too many hashing collisions. """ REDUCERS[datatype] = func def make(columns): """Build an hashable tuple subclass.""" # When all columns are hashable, pass the input tuple through as is. if all(column.datatype in FUNDAMENTAL for column in columns): return lambda x: x datatypes = ', '.join(types.name(column.datatype) for column in columns) # Code generation inspired by standard library ``dataclasses.py``. parts = [] locals = {} for i, column in enumerate(columns): if column.datatype in FUNDAMENTAL: parts.append(f'self[{i}]') elif column.datatype is dict: parts.append(f'*self[{i}].keys(), *self[{i}].values()') elif column.datatype is set: parts.append(f'*self[{i}]') else: func = REDUCERS.get(column.datatype, pickle.dumps) fname = f'func{i}' locals[fname] = func parts.append(f'{fname}(self[{i}])') objs = ', '.join(parts) names = ', '.join(locals.keys()) code = textwrap.dedent(f''' def create({names}): def __hash__(self): return hash(({objs})) return __hash__ ''') clsname = f'Hashable[{datatypes}]' ns = {} exec(code, globals(), ns) func = ns['create'](**locals) func.__qualname__ = f'{clsname}.{func.__name__}' members = dict(tuple.__dict__) members['__hash__'] = func return type(clsname, (tuple,), members) beanquery-0.2.0/beanquery/hashable_test.py000066400000000000000000000026721477033713200206750ustar00rootroot00000000000000import dataclasses import unittest from beanquery import hashable from beanquery.cursor import Column class TestHashable(unittest.TestCase): def test_fundamental(self): columns = (Column('b', bool), Column('i', int), Column('s', str)) wrap = hashable.make(columns) obj = (True, 42, 'universe') self.assertIs(wrap(obj), obj) hash(obj) def test_dict(self): columns = (Column('b', bool), Column('d', dict)) wrap = hashable.make(columns) obja = (True, {'answer': 42}) a = hash(wrap(obja)) objb = (True, {'answer': 42}) b = hash(wrap(objb)) self.assertIsNot(obja, objb) self.assertEqual(a, b) objc = (False, {'answer': 42}) c = hash(wrap(objc)) self.assertNotEqual(a, c) objd = (True, {'answer': 43}) d = hash(wrap(objd)) self.assertNotEqual(a, d) def test_registered(self): @dataclasses.dataclass class Foo: xid: int meta: dict hashable.register(Foo, lambda obj: obj.xid) columns = (Column('b', bool), Column('foo', Foo)) wrap = hashable.make(columns) obja = (True, Foo(1, {'test': 1})) a = hash(wrap(obja)) objb = (True, Foo(1, {'test': 2})) b = hash(wrap(objb)) self.assertEqual(a, b) objc = (True, Foo(2, {'test': 2})) c = hash(wrap(objc)) self.assertNotEqual(a, c) beanquery-0.2.0/beanquery/numberify.py000066400000000000000000000203241477033713200200610ustar00rootroot00000000000000"""Code to split table columns containing amounts and inventories into number columns. For example, given a column with this content: ----- amount ------ 101.23 USD 200 JPY 99.23 USD 38.34 USD, 100 JPY We can convert this into two columns and remove the currencies: -amount (USD)- -amount (JPY)- 101.23 200 99.23 38.34 100 The point is that the columns should be typed as numbers to make this importable into a spreadsheet and able to be processed. Notes: * This handles the Amount, Position and Inventory datatypes. There is code to automatically recognize columns containing such types from a table of strings and convert such columns to their corresponding guessed data types. * The per-currency columns are ordered in decreasing order of the number of instances of numbers seen for each currency. So if the most numbers you have in a column are USD, then the USD column renders first. * Cost basis specifications should be unmodified and reported to a dedicated extra column, like this: ----- amount ------ 1 AAPL {21.23 USD} We can convert this into two columns and remove the currencies: -amount (AAPL)- -Cost basis- 1 {21.23 USD} (Eventually we might support the conversion of cost amounts as well, but they may contain other information, such as a label or a date, so for now we don't convert them. I'm not sure there's a good practical use case in doing that yet.) * We may provide some options to break out only some of the currencies into columns, in order to handle the case where an inventory contains a large number of currencies and we want to only operate on a restricted set of operating currencies. * If you provide a DisplayFormatter object to the numberification routine, they quantize each column according to their currency's precision. It is recommended that you do that. """ __copyright__ = "Copyright (C) 2015-2017 Martin Blais" __license__ = "GNU GPLv2" import collections from decimal import Decimal from beancount.core import amount from beancount.core import position from beancount.core import inventory from .cursor import Column def numberify_results(columns, drows, dformat=None): """Number rows containing Amount, Position or Inventory types. Args: result_types: A list of items describing the names and data types of the items in each column. result_rows: A list of ResultRow instances. dformat: An optional DisplayFormatter. If set, quantize the numbers by their currency-specific precision when converting the Amount's, Position's or Inventory'es.. Returns: A pair of modified (result_types, result_rows) with converted datatypes. """ # Build an array of converters. converters = [] for index, column in enumerate(columns): convert_col_fun = CONVERTING_TYPES.get(column.datatype) if convert_col_fun is None: converters.append(IdentityConverter(column.name, column.datatype, index)) else: col_converters = convert_col_fun(column.name, drows, index) converters.extend(col_converters) # Derive the output types from the expected outputs from the converters # themselves. otypes = tuple(Column(c.name, c.dtype) for c in converters) # Convert the input rows by processing them through the converters. orows = [] for drow in drows: orow = [] for converter in converters: orow.append(converter(drow, dformat)) orows.append(orow) return otypes, orows class IdentityConverter: """A converter that simply copies its column.""" def __init__(self, name, dtype, index): self.name = name self.dtype = dtype self.index = index def __call__(self, drow, _): return drow[self.index] class AmountConverter: """A converter that extracts the number of an amount for a specific currency.""" dtype = Decimal def __init__(self, name, index, currency): self.name = name self.index = index self.currency = currency def __call__(self, drow, dformat): vamount = drow[self.index] if vamount and vamount.currency == self.currency: number = vamount.number if dformat: number = dformat.quantize(number, self.currency) else: number = None return number def convert_col_Amount(name, drows, index): """Create converters for a column of type Amount. Args: name: A string, the column name. drows: The table of objects. index: The column number. Returns: A list of Converter instances, one for each of the currency types found. """ currency_map = collections.defaultdict(int) for drow in drows: vamount = drow[index] if vamount and vamount.currency: currency_map[vamount.currency] += 1 return [AmountConverter('{} ({})'.format(name, currency), index, currency) for currency, _ in sorted(currency_map.items(), key=lambda item: (item[1], item[0]), reverse=True)] class PositionConverter: """A converter that extracts the number of a position for a specific currency.""" dtype = Decimal def __init__(self, name, index, currency): self.name = name self.index = index self.currency = currency def __call__(self, drow, dformat): pos = drow[self.index] if pos and pos.units.currency == self.currency: number = pos.units.number if dformat: number = dformat.quantize(pos.units.number, self.currency) else: number = None return number def convert_col_Position(name, drows, index): """Create converters for a column of type Position. Args: name: A string, the column name. drows: The table of objects. index: The column number. Returns: A list of Converter instances, one for each of the currency types found. """ currency_map = collections.defaultdict(int) for drow in drows: pos = drow[index] if pos and pos.units.currency: currency_map[pos.units.currency] += 1 return [PositionConverter('{} ({})'.format(name, currency), index, currency) for currency, _ in sorted(currency_map.items(), key=lambda item: (item[1], item[0]), reverse=True)] class InventoryConverter: """A converter that extracts the number of a inventory for a specific currency. If there are multiple lots we aggregate by currency.""" dtype = Decimal def __init__(self, name, index, currency): self.name = name self.index = index self.currency = currency def __call__(self, drow, dformat): inv = drow[self.index] # FIXME:: get_currency_units() returns ZERO and not None when the value # isn't present. This should be fixed to distinguish between the two. number = inv.get_currency_units(self.currency).number if number and dformat: number = dformat.quantize(number, self.currency) return number or None def convert_col_Inventory(name, drows, index): """Create converters for a column of type Inventory. Args: name: A string, the column name. drows: The table of objects. index: The column number. Returns: A list of Converter instances, one for each of the currency types found. """ currency_map = collections.defaultdict(int) for drow in drows: inv = drow[index] for currency in inv.currencies(): currency_map[currency] += 1 return [InventoryConverter('{} ({})'.format(name, currency), index, currency) for currency, _ in sorted(currency_map.items(), key=lambda item: (item[1], item[0]), reverse=True)] # A mapping of data types to their converter factory. CONVERTING_TYPES = { amount.Amount : convert_col_Amount, position.Position : convert_col_Position, inventory.Inventory : convert_col_Inventory, } beanquery-0.2.0/beanquery/numberify_test.py000066400000000000000000000131231477033713200211170ustar00rootroot00000000000000__copyright__ = "Copyright (C) 2015-2017 Martin Blais" __license__ = "GNU GPLv2" import datetime import unittest from decimal import Decimal from beancount.core.number import D from beancount.core.amount import A from beancount.core import amount from beancount.core import position from beancount.core import inventory from beancount.core import display_context from beanquery import numberify from beanquery.cursor import Column class TestNumerifySimple(unittest.TestCase): input_amounts = ["24.17 CAD", "-77.02 CAD", "11.39 CAD", "800.00 USD", "41.17 CAD", "950.00 USD", "110 JPY", "-947.00 USD"] expected_types = (('pos (CAD)', Decimal), ('pos (USD)', Decimal), ('pos (JPY)', Decimal)) expected_rows = [[D('24.17'), None, None], [D('-77.02'), None, None], [D('11.39'), None, None], [None, D('800.00'), None], [D('41.17'), None, None], [None, D('950.00'), None], [None, None, D('110')], [None, D('-947.00'), None]] def test_amount(self): itypes = (Column('pos', amount.Amount), ) irows = [(A(string),) for string in self.input_amounts] atypes, arows = numberify.numberify_results(itypes, irows) self.assertEqual(self.expected_types, atypes) self.assertEqual(self.expected_rows, arows) def test_position(self): itypes = (Column('pos', position.Position), ) irows = [(position.from_string(string),) for string in self.input_amounts] atypes, arows = numberify.numberify_results(itypes, irows) self.assertEqual(self.expected_types, atypes) self.assertEqual(self.expected_rows, arows) def test_inventory(self): itypes = (Column('pos', inventory.Inventory), ) irows = [(inventory.from_string(string),) for string in self.input_amounts] atypes, arows = numberify.numberify_results(itypes, irows) self.assertEqual(self.expected_types, atypes) self.assertEqual(self.expected_rows, arows) class TestNumerifyIdentity(unittest.TestCase): def test_identity(self): itypes = (Column('date', datetime.date), Column('name', str), Column('count', int), ) irows = [[datetime.date(2015, 9, 8), 'Testing', 3]] atypes, arows = numberify.numberify_results(itypes, irows) self.assertEqual(itypes, atypes) self.assertEqual(irows, arows) class TestNumerifyInventory(unittest.TestCase): def test_inventory(self): itypes = (Column('balance', inventory.Inventory), ) irows = [[inventory.from_string('10 HOOL {23.00 USD}')], [inventory.from_string('2.11 USD, 3.44 CAD')], [inventory.from_string('-2 HOOL {24.00 USD}, 5.66 CAD')]] atypes, arows = numberify.numberify_results(itypes, irows) self.assertEqual((('balance (HOOL)', Decimal), ('balance (CAD)', Decimal), ('balance (USD)', Decimal), ), atypes) self.assertEqual([[D('10'), None, None], [None, D('3.44'), D('2.11')], [D('-2'), D('5.66'), None]], arows) class TestNumerifyPrecision(unittest.TestCase): def test_precision(self): # Some display context. dcontext = display_context.DisplayContext() dcontext.update(D('111'), 'JPY') dcontext.update(D('1.111'), 'RGAGX') dcontext.update(D('1.11'), 'USD') dformat = dcontext.build() # Input data. itypes = (Column('number', Decimal), Column('amount', amount.Amount), Column('position', position.Position), Column('inventory', inventory.Inventory)) irows = [[D(amt.split()[0]), A(amt), position.from_string(amt), inventory.from_string(amt)] for amt in ['123.45678909876 JPY', '1.67321232123 RGAGX', '5.67345434543 USD']] # First check with no explicit quantization. atypes, arows = numberify.numberify_results(itypes, irows) erows = [[D('123.45678909876'), None, None, D('123.45678909876'), None, None, D('123.45678909876'), None, None, D('123.45678909876')], [D('1.67321232123'), None, D('1.67321232123'), None, None, D('1.67321232123'), None, None, D('1.67321232123'), None], [D('5.67345434543'), D('5.67345434543'), None, None, D('5.67345434543'), None, None, D('5.67345434543'), None, None]] self.assertEqual(erows, arows) # Then compare with quantization. atypes, arows = numberify.numberify_results(itypes, irows, dformat) erows = [[D('123.45678909876'), None, None, D('123'), None, None, D('123'), None, None, D('123')], [D('1.67321232123'), None, D('1.673'), None, None, D('1.673'), None, None, D('1.673'), None], [D('5.67345434543'), D('5.67'), None, None, D('5.67'), None, None, D('5.67'), None, None]] self.assertEqual(erows, arows) if __name__ == '__main__': unittest.main() beanquery-0.2.0/beanquery/parser/000077500000000000000000000000001477033713200170025ustar00rootroot00000000000000beanquery-0.2.0/beanquery/parser/__init__.py000066400000000000000000000031531477033713200211150ustar00rootroot00000000000000import datetime import decimal import tatsu from ..errors import ProgrammingError from .parser import BQLParser from . import ast class BQLSemantics: def set_context(self, ctx): self._ctx = ctx def null(self, value): return None def integer(self, value): return int(value) def decimal(self, value): return decimal.Decimal(value) def date(self, value): return datetime.date.fromisoformat(value) def string(self, value): return value[1:-1] def boolean(self, value): return value == 'TRUE' def unquoted_identifier(self, value): return value.lower() def quoted_identifier(self, value): return value.replace('""', '"') def asterisk(self, value): return ast.Asterisk() def list(self, value): return list(value) def ordering(self, value): return ast.Ordering[value or 'ASC'] def _default(self, value, typename=None): if typename is not None: func = getattr(ast, typename) return func(**{name.rstrip('_'): value for name, value in value.items()}) return value class ParseError(ProgrammingError): def __init__(self, parseinfo): super().__init__('syntax error') self.parseinfo = parseinfo def parse(text): try: return BQLParser().parse(text, semantics=BQLSemantics()) except tatsu.exceptions.ParseError as exc: line = exc.tokenizer.line_info(exc.pos).line parseinfo = tatsu.infos.ParseInfo(exc.tokenizer, exc.item, exc.pos, exc.pos + 1, line, []) raise ParseError(parseinfo) from exc beanquery-0.2.0/beanquery/parser/ast.py000066400000000000000000000174071477033713200201540ustar00rootroot00000000000000from __future__ import annotations import dataclasses import datetime import enum import sys import textwrap import typing if typing.TYPE_CHECKING: from typing import Any, Optional, Union def _indent(text): return textwrap.indent(text, ' ') def _fields(node): for field in dataclasses.fields(node): if field.repr: yield field.name, getattr(node, field.name) def tosexp(node): if isinstance(node, Node): return f'({node.__class__.__name__.lower()}\n' + _indent( '\n'.join(f'{name.replace("_", "-")}: {tosexp(value)}' for name, value in _fields(node) if value is not None) + ')') if isinstance(node, list): return '(\n' + _indent('\n'.join(tosexp(i) for i in node)) + ')' if isinstance(node, enum.Enum): return node.name.lower() return repr(node) def walk(node): if isinstance(node, Node): for name, child in _fields(node): yield from walk(child) yield node if isinstance(node, list): for child in node: yield from walk(child) class Node: """Base class for BQL AST nodes.""" __slots__ = () parseinfo = None @property def text(self): if not self.parseinfo: return None text = self.parseinfo.tokenizer.text return text[self.parseinfo.pos:self.parseinfo.endpos] def tosexp(self): return tosexp(self) def walk(self): return walk(self) def node(name, fields): """Manufacture an AST node class.""" return dataclasses.make_dataclass( name, [*fields.split(), ('parseinfo', None, dataclasses.field(default=None, compare=False, repr=False))], bases=(Node,), **({'slots': True} if sys.version_info[:2] >= (3, 10) else {})) # A 'select' query action. # # Attributes: # targets: Either a single 'Asterisk' instance of a list of 'Target' # instances. # from_clause: An instance of 'From', or None if absent. # where_clause: A root expression node, or None if absent. # group_by: An instance of 'GroupBy', or None if absent. # order_by: An instance of 'OrderBy', or None if absent. # pivot_by: An instance of 'PivotBy', or None if absent. # limit: An integer, or None is absent. # distinct: A boolean value (True), or None if absent. Select = node('Select', 'targets from_clause where_clause group_by order_by pivot_by limit distinct') # A select query that produces final balances for accounts. # This is equivalent to # # SELECT account, sum(position) # FROM ... # WHERE ... # GROUP BY account # # Attributes: # summary_func: A method on an inventory to call on the position column. # May be to extract units, value at cost, etc. # from_clause: An instance of 'From', or None if absent. Balances = node('Balances', 'summary_func from_clause where_clause') # A select query that produces a journal of postings. # This is equivalent to # # SELECT date, flag, payee, narration, ... FROM # WHERE account = # # Attributes: # account: A string, the name of the account to restrict to. # summary_func: A method on an inventory to call on the position column. # May be to extract units, value at cost, etc. # from_clause: An instance of 'From', or None if absent. Journal = node('Journal', 'account summary_func from_clause') # A query that will simply print the selected entries in Beancount format. # # Attributes: # from_clause: An instance of 'From', or None if absent. Print = node('Print', 'from_clause') # A parsed SELECT column or target. # # Attributes: # expression: A tree of expression nodes from the parser. # name: A string, the given name of the target (given by "AS "). Target = node('Target', 'expression name') # A placeholder in SELECT * or COUNT(*) constructs. Asterisk = node('Asterisk', '') # A FROM clause. # # Attributes: # expression: A tree of expression nodes from the parser. # close: A CLOSE clause, either None if absent, a boolean if the clause # was present by no date was provided, or a datetime.date instance if # a date was provided. @dataclasses.dataclass(**({'slots': True} if sys.version_info[:2] >= (3, 10) else {})) class From(Node): expression: Optional[Node] = None open: Optional[datetime.date] = None close: Optional[Union[datetime.date, bool]] = None clear: Optional[bool] = None parseinfo: Any = dataclasses.field(default=None, compare=False, repr=False) # A GROUP BY clause. # # Attributes: # columns: A list of group-by expressions, simple Column() or otherwise. # having: An expression tree for the optional HAVING clause, or None. GroupBy = node('GroupBy', 'columns having') # An ORDER BY clause. # # Attributes: # column: order-by expression, simple Column() or otherwise. # ordering: The sort order as an Ordering enum value. OrderBy = node('OrderBy', 'column ordering') class Ordering(enum.IntEnum): # The enum values are chosen in this way to be able to use them # directly as the reverse parameter to the list sort() method. ASC = 0 DESC = 1 def __repr__(self): return f"{self.__class__.__name__}.{self.name}" # An PIVOT BY clause. # # Attributes: # columns: A list of group-by expressions, simple Column() or otherwise. PivotBy = node('PivotBy', 'columns') # A reference to a table. # # Attributes: # name: The table name. Table = node('Table', 'name') # A reference to a column. # # Attributes: # name: A string, the name of the column to access. Column = node('Column', 'name') # A function call. # # Attributes: # fname: A string, the name of the function. # operands: A list of other expressions, the arguments of the function to # evaluate. This is possibly an empty list. Function = node('Function', 'fname operands') Attribute = node('Attribute', 'operand name') Subscript = node('Subscript', 'operand key') # A constant node. # # Attributes: # value: The constant value this represents. Constant = node('Constant', 'value') # A query parameter placeholder. # # Attributes: # name: The placeholder name Placeholder = node('Placeholder', 'name') # Base class for unary operators. # # Attributes: # operand: An expression, the operand of the operator. UnaryOp = node('UnaryOp', 'operand') # Base class for binary operators. # # Attributes: # left: An expression, the left operand. # right: An expression, the right operand. BinaryOp = node('BinaryOp', 'left right') # Base class for boolean operators. BoolOp = node('BoolOp', 'args') # Between Between = node('Between', 'operand lower upper') # Negation operator. class Not(UnaryOp): __slots__ = () class IsNull(UnaryOp): __slots__ = () class IsNotNull(UnaryOp): __slots__ = () # Boolean operators. class And(BoolOp): __slots__ = () class Or(BoolOp): __slots__ = () # Equality and inequality comparison operators. class Equal(BinaryOp): __slots__ = () class NotEqual(BinaryOp): __slots__ = () class Greater(BinaryOp): __slots__ = () class GreaterEq(BinaryOp): __slots__ = () class Less(BinaryOp): __slots__ = () class LessEq(BinaryOp): __slots__ = () # Regular expression match operator. class Match(BinaryOp): __slots__ = () class NotMatch(BinaryOp): __slots__ = () class Matches(BinaryOp): __slots__ = () # Membership operators. class In(BinaryOp): __slots__ = () class NotIn(BinaryOp): __slots__ = () # Arithmetic operators. class Neg(UnaryOp): __slots__ = () class Mul(BinaryOp): __slots__ = () class Div(BinaryOp): __slots__ = () class Mod(BinaryOp): __slots__ = () class Add(BinaryOp): __slots__ = () class Sub(BinaryOp): __slots__ = () Any = node('Any', 'left op right') All = node('All', 'left op right') CreateTable = node('CreateTable', 'name columns using query') Insert = node('Insert', 'table columns values') beanquery-0.2.0/beanquery/parser/bql.ebnf000066400000000000000000000137671477033713200204320ustar00rootroot00000000000000@@grammar :: BQL @@parseinfo :: True @@ignorecase :: True @@keyword :: 'AND' 'AS' 'ASC' 'BY' 'DESC' 'DISTINCT' 'FALSE' 'FROM' 'GROUP' 'HAVING' 'IN' 'IS' 'LIMIT' 'NOT' 'OR' 'ORDER' 'PIVOT' 'SELECT' 'TRUE' 'WHERE' @@keyword :: 'CREATE' 'TABLE' 'USING' 'INSERT' 'INTO' @@keyword :: 'BALANCES' 'JOURNAL' 'PRINT' @@comments :: /(\/\*([^*]|[\r\n]|(\*+([^*\/]|[\r\n])))*\*+\/)/ @@eol_comments :: /\;[^\n]*?$/ bql = @:statement [';'] $ ; statement = | select | balances | journal | print | create_table | insert ; select::Select = 'SELECT' ['DISTINCT' distinct:`True`] targets:(','.{ target }+ | asterisk) ['FROM' from_clause:(_table | subselect | from)] ['WHERE' where_clause:expression] ['GROUP' 'BY' group_by:groupby] ['ORDER' 'BY' order_by:','.{order}+] ['PIVOT' 'BY' pivot_by:pivotby] ['LIMIT' limit:integer] ; subselect = '(' @:select ')' ; from::From = | 'OPEN' ~ 'ON' open:date ['CLOSE' ('ON' close:date | {} close:`True`)] ['CLEAR' clear:`True`] | 'CLOSE' ~ ('ON' close:date | {} close:`True`) ['CLEAR' clear:`True`] | 'CLEAR' ~ clear:`True` | expression:expression ['OPEN' 'ON' open:date] ['CLOSE' ('ON' close:date | {} close:`True`)] ['CLEAR' clear:`True`] ; _table::Table = | name:/#([a-zA-Z_][a-zA-Z0-9_]*)?/ | name:quoted_identifier ; table::Table = name:identifier ; groupby::GroupBy = columns:','.{ (integer | expression) }+ ['HAVING' having:expression] ; order::OrderBy = column:(integer | expression) ordering:ordering ; ordering = ['DESC' | 'ASC'] ; pivotby::PivotBy = columns+:(integer | column) ',' columns+:(integer | column) ; target::Target = expression:expression ['AS' name:identifier] ; expression = | disjunction | conjunction ; disjunction = | or | conjunction ; or::Or::BoolOp = args+:conjunction { 'OR' args+:conjunction }+ ; conjunction = | and | inversion ; and::And::BoolOp = args+:inversion { 'AND' args+:inversion }+ ; inversion = | not | comparison ; not::Not::UnaryOp = 'NOT' operand:inversion ; comparison = | any | all | lt | lte | gt | gte | eq | neq | in | notin | match | notmatch | matches | isnull | isnotnull | between | sum ; any::Any = left:sum op:op 'any' '(' right:expression ')' ; all::All = left:sum op:op 'all' '(' right:expression ')' ; op = | '<' | '<=' | '>' | '>=' | '=' | '!=' | '~' | '!~' | '?~' ; lt::Less::BinaryOp = left:sum '<' right:sum ; lte::LessEq::BinaryOp = left:sum '<=' right:sum ; gt::Greater::BinaryOp = left:sum '>' right:sum ; gte::GreaterEq::BinaryOp = left:sum '>=' right:sum ; eq::Equal::BinaryOp = left:sum '=' right:sum ; neq::NotEqual::BinaryOp = left:sum '!=' right:sum ; in::In::BinaryOp = left:sum 'IN' right:sum ; notin::NotIn::BinaryOp = left:sum 'NOT' 'IN' right:sum ; match::Match::BinaryOp = left:sum '~' right:sum ; notmatch::NotMatch::BinaryOp = left:sum '!~' right:sum ; matches::Matches::BinaryOp = left:sum '?~' right:sum ; isnull::IsNull::UnaryOp = operand:sum 'IS' 'NULL' ; isnotnull::IsNotNull::UnaryOp = operand:sum 'IS' 'NOT' 'NULL' ; between::Between = operand:sum 'BETWEEN' lower:sum 'AND' upper:sum ; sum = | add | sub | term ; add::Add::BinaryOp = left:sum '+' ~ right:term ; sub::Sub::BinaryOp = left:sum '-' ~ right:term ; term = | mul | div | mod | factor ; mul::Mul::BinaryOp = left:term '*' ~ right:factor ; div::Div::BinaryOp = left:term '/' ~ right:factor ; mod::Mod::BinaryOp = left:term '%' ~ right:factor ; factor = | unary | '(' @:expression ')' ; unary = | uplus | uminus | primary ; uplus = '+' @:atom ; uminus::Neg::UnaryOp = '-' operand:factor ; primary = | attribute | subscript | atom ; attribute::Attribute = operand:primary '.' name:identifier ; subscript::Subscript = operand:primary '[' key:string ']' ; atom = | select | function | constant | column | placeholder ; placeholder::Placeholder = | '%s' name:`` | '%(' name:identifier ')s' ; function::Function = | fname:identifier '(' operands:','.{ expression } ')' | fname:identifier '(' operands+:asterisk ')' ; column::Column = name:identifier ; literal = | date | decimal | integer | string | null | boolean ; constant::Constant = value:(literal | list) ; list = '(' &( literal ',') @:','.{ (literal | ()) }+ ')' ; identifier = | unquoted_identifier | quoted_identifier ; @name unquoted_identifier = /[a-zA-Z_][a-zA-Z0-9_]*/ ; quoted_identifier = /\"((?:[^\"]|\"\")+)\"/ ; asterisk = '*' ; string = /\"[^\"]*\"|\'(?:[^\']|\'\')*\'/ ; boolean = 'TRUE' | 'FALSE' ; null = 'NULL' ; integer = /[0-9]+/ ; decimal = /[0-9]+\.[0-9]*|[0-9]*\.[0-9]+/ ; date = /[0-9]{4}-[0-9]{2}-[0-9]{2}/ ; balances::Balances = 'BALANCES' ['AT' summary_func:identifier] ['FROM' from_clause:from] ['WHERE' where_clause:expression] ; journal::Journal = 'JOURNAL' [account:string] ['AT' summary_func:identifier] ['FROM' from_clause:from] ; print::Print = 'PRINT' ['FROM' from_clause:from] ; create_table::CreateTable = 'CREATE' 'TABLE' ~ name:identifier ( | '(' columns:','.{( identifier identifier )} ')' ['USING' using:string] | 'USING' using:string | 'AS' query:select ) ; insert::Insert = 'INSERT' 'INTO' ~ table:table ['(' columns:','.{column} ')'] 'VALUES' '(' values:','.{expression} ')' ; beanquery-0.2.0/beanquery/parser/parser.py000066400000000000000000001132441477033713200206550ustar00rootroot00000000000000#!/usr/bin/env python3 # WARNING: CAVEAT UTILITOR # # This file was automatically generated by TatSu. # # https://pypi.python.org/pypi/tatsu/ # # Any changes you make to it will be overwritten the next time # the file is generated. # ruff: noqa: C405, COM812, I001, F401, PLR1702, PLC2801, SIM117 from __future__ import annotations import sys from pathlib import Path from tatsu.buffering import Buffer from tatsu.parsing import Parser from tatsu.parsing import tatsumasu from tatsu.parsing import leftrec, nomemo, isname from tatsu.infos import ParserConfig from tatsu.util import re, generic_main KEYWORDS: set[str] = { 'AND', 'AS', 'ASC', 'BY', 'DESC', 'DISTINCT', 'FALSE', 'FROM', 'GROUP', 'HAVING', 'IN', 'IS', 'LIMIT', 'NOT', 'OR', 'ORDER', 'PIVOT', 'SELECT', 'TRUE', 'WHERE', 'CREATE', 'TABLE', 'USING', 'INSERT', 'INTO', 'BALANCES', 'JOURNAL', 'PRINT', } class BQLBuffer(Buffer): def __init__(self, text, /, config: ParserConfig | None = None, **settings): config = ParserConfig.new( config, owner=self, whitespace=None, nameguard=None, ignorecase=True, namechars='', parseinfo=True, comments='(\\/\\*([^*]|[\\r\\n]|(\\*+([^*\\/]|[\\r\\n])))*\\*+\\/)', eol_comments='\\;[^\\n]*?$', keywords=KEYWORDS, start='bql', ) config = config.replace(**settings) super().__init__(text, config=config) class BQLParser(Parser): def __init__(self, /, config: ParserConfig | None = None, **settings): config = ParserConfig.new( config, owner=self, whitespace=None, nameguard=None, ignorecase=True, namechars='', parseinfo=True, comments='(\\/\\*([^*]|[\\r\\n]|(\\*+([^*\\/]|[\\r\\n])))*\\*+\\/)', eol_comments='\\;[^\\n]*?$', keywords=KEYWORDS, start='bql', ) config = config.replace(**settings) super().__init__(config=config) @tatsumasu() def _bql_(self): self._statement_() self.name_last_node('@') with self._optional(): self._token(';') self._check_eof() @tatsumasu() def _statement_(self): with self._choice(): with self._option(): self._select_() with self._option(): self._balances_() with self._option(): self._journal_() with self._option(): self._print_() with self._option(): self._create_table_() with self._option(): self._insert_() self._error( 'expecting one of: ' "'BALANCES' 'CREATE' 'INSERT' 'JOURNAL'" "'PRINT' 'SELECT' " ' ' ' ' ) @tatsumasu('Attribute') @nomemo def _attribute_(self): self._primary_() self.name_last_node('operand') self._token('.') self._identifier_() self.name_last_node('name') self._define(['name', 'operand'], []) @tatsumasu('Subscript') @nomemo def _subscript_(self): self._primary_() self.name_last_node('operand') self._token('[') self._string_() self.name_last_node('key') self._token(']') self._define(['key', 'operand'], []) @tatsumasu() def _atom_(self): with self._choice(): with self._option(): self._select_() with self._option(): self._function_() with self._option(): self._constant_() with self._option(): self._column_() with self._option(): self._placeholder_() self._error( 'expecting one of: ' "'%(' '%s' 'SELECT' " ' ' ' ' '