coalib.bearlib.languages.documentation.extract_documentation() - Code Metrics - Inspection of "Makman2/doc format bear8" - coala-analyzer/coala - Measure and Improve Code Quality continuously with Scrutinizer

Failed Conditions

Pull Request — master (#1139)

by Mischa

created 2015-12-16 20:01 UTC

coalib.bearlib.languages.documentation.extract_documentation() B

↳ Parent: Project

Complexity

Conditions

Size

Total Lines

Duplication

Lines	0
Ratio	0 %

Metric	Value
cc	1
dl	0
loc	30
rs	8.8571

import re

from coalib.bearlib.languages.documentation.DocstyleDefinition import (
    DocstyleDefinition)
from coalib.bearlib.languages.documentation.DocumentationComment import (
    DocumentationComment)
from coalib.results.TextRange import TextRange


def _extract_doccomment_simple(content, line, column, markers):
    """
    Extract a documentation that starts at given beginning with simple layout.

    The property of the simple layout is that there's no each-line marker. This
    applies e.g. for python docstrings.

    :param content: Presplitted lines of the source-code-string.
    :param line:    Line where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param column:  Column where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param markers: The documentation identifying markers.
    :return:        If the comment matched layout a triple with end-of-comment
                    line, column and the extracted documentation. If not
                    matched, returns None.
    """
    align_column = column - len(markers[0])

    pos = content[line].find(markers[2], column)
    if pos == -1:
        doccomment = content[line][column:]
        line += 1

        while line < len(content):
            pos = content[line].find(markers[2])
            if pos == -1:
                doccomment += content[line][align_column:]
            else:
                doccomment += content[line][align_column:pos]
                return line, pos + len(markers[2]), doccomment

            line += 1

        return None
    else:
        return line, pos + len(markers[2]), content[line][column:pos]


def _extract_doccomment_continuous(content, line, column, markers):
    """
    Extract a documentation that starts at given beginning with continuous
    layout.

    The property of the continuous layout is that the each-line-marker and the
    end-marker do equal. Documentation is extracted until no further marker is
    found. Applies e.g. for doxygen style python documentation:

    ```
    ## main
    #
    #  detailed
    ```

    :param content: Presplitted lines of the source-code-string.
    :param line:    Line where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param column:  Column where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param markers: The documentation identifying markers.
    :return:        If the comment matched layout a triple with end-of-comment
                    line, column and the extracted documentation. If not
                    matched, returns None.
    """
    marker_len = len(markers[1])

    doccomment = content[line][column:]
    line += 1
    while line < len(content):
        pos = content[line].find(markers[1])
        if pos == -1:
            return line, 0, doccomment
        else:
            doccomment += content[line][pos+marker_len:]

        line += 1

    if content[line - 1][-1] == "\n":
        column = 0
    else:
        # This case can appear on end-of-document without a `\n`.
        line -= 1
        column = len(content[line])

    return line, column, doccomment


def _extract_doccomment_standard(content, line, column, markers):
    """
    Extract a documentation that starts at given beginning with standard
    layout.

    The standard layout applies e.g. for C doxygen-style documentation:

    ```
    /**
     * documentation
     */
    ```

    :param content: Presplitted lines of the source-code-string.
    :param line:    Line where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param column:  Column where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param markers: The documentation identifying markers.
    :return:        If the comment matched layout a triple with end-of-comment
                    line, column and the extracted documentation. If not
                    matched, returns None.
    """
    pos = content[line].find(markers[2], column)
    if pos != -1:
        return line, pos + len(markers[2]), content[line][column:pos]
    else:
        doccomment = content[line][column:]
        line += 1

        while line < len(content):
            pos = content[line].find(markers[2])
            each_line_pos = content[line].find(markers[1])

            if pos == -1:
                if each_line_pos == -1:
                    # If the first text occurrence is not the each-line marker
                    # now we violate the doc-comment layout.
                    return None
                doccomment += content[line][each_line_pos+len(markers[1]):]
            else:
                # If no each-line marker found or it's located past the end
                # marker: extract no further and end the doc-comment.
                if each_line_pos != -1 and each_line_pos + 1 < pos:
                    doccomment += (
                        content[line][each_line_pos+len(markers[1]):pos])

                return line, pos + len(markers[2]), doccomment

            line += 1

        return None


def _extract_doccomment(content, line, column, markers):
    """
    Delegates depending on the given markers to the right extraction method.

    :param content: Presplitted lines of the source-code-string.
    :param line:    Line where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param column:  Column where the documentation comment starts (behind the
                    start marker). Zero-based.
    :param markers: The documentation identifying markers.
    :return:        If the comment matched layout a triple with end-of-comment
                    line, column and the extracted documentation. If not
                    matched, returns None.
    """
    if markers[1] == "":
        # Extract and align to start marker.
        return _extract_doccomment_simple(content, line, column, markers)
    elif markers[1] == markers[2]:
        # Search for the each-line marker until it runs out.
        return _extract_doccomment_continuous(content, line, column, markers)
    else:
        return _extract_doccomment_standard(content, line, column, markers)


def _compile_multi_match_regex(strings):
    """
    Compiles a regex object that matches each of the given strings.

    :param strings: The strings to match.
    :return:        A regex object.
    """
    return re.compile("|".join(re.escape(s) for s in strings))


def _extract_doccomment_from_line(content, line, column, regex, marker_dict):
    begin_match = regex.search(content[line], column)
    if begin_match:
        column = begin_match.end()
        for marker in marker_dict[begin_match.group()]:
            doccomment = _extract_doccomment(content, line, column, marker)
            if doccomment is not None:
                end_line, end_column, documentation = doccomment


                rng = TextRange.from_values(line + 1,
                                            begin_match.start() + 1,
                                            end_line + 1,
                                            end_column + 1)
                doc = DocumentationComment(documentation, marker, rng)

                return end_line, end_column, doc

    return line + 1, 0, None


def extract_documentation_with_markers(content, markers):
    """
    Extracts all documentation texts inside the given source-code-string.

    :param content: The source-code-string where to extract documentation from.
                    Needs to be a list or tuple where each string item is a
                    single line (including ending whitespaces like `\\n`).
    :param markers: The list/tuple of marker-sets that identify a
                    documentation-comment. Low-index markers have higher
                    priority than high-index markers.
    :return:        An iterator returning each DocumentationComment found in
                    the content.
    """
    # Prepare marker-tuple dict that maps a begin pattern to the corresponding
    # marker_set(s). This makes it faster to retrieve a marker-set from a
    # begin sequence we initially want to search for in source code. Then
    # the possible found documentation match is processed further with the
    # rest markers.
    marker_dict = {}
    for marker_set in markers:
        if marker_set[0] not in marker_dict:
            marker_dict[marker_set[0]] = [marker_set]
        else:
            marker_dict[marker_set[0]].append(marker_set)

    # Using regexes to perform a variable match is faster than finding each
    # substring with `str.find()` choosing the lowest match.
    begin_regex = _compile_multi_match_regex(
        marker_set[0] for marker_set in markers)

    line = 0
    column = 0
    while line < len(content):
        line, column, doc = _extract_doccomment_from_line(content,
                                                          line,
                                                          column,
                                                          begin_regex,
                                                          marker_dict)
        if doc:
            yield doc


def extract_documentation(content, language, docstyle):
    """
    Extracts all documentation texts inside the given source-code-string using
    the coala docstyle definition files.

    The documentation texts are sorted by their order appearing in `content`.

    For more information about how documentation comments are identified and
    extracted, see DocstyleDefinition.doctypes enumeration.

    :param content:            The source-code-string where to extract
                               documentation from. Needs to be a list or tuple
                               where each string item is a single line
                               (including ending whitespaces like `\\n`).
    :param language:           The programming language used.
    :param docstyle:           The documentation style/tool used
                               (e.g. doxygen).
    :raises FileNotFoundError: Raised when the docstyle definition file was not
                               found. This is a compatability exception from
                               `coalib.misc.Compatability` module.
    :raises KeyError:          Raised when the given language is not defined in
                               given docstyle.
    :raises ValueError:        Raised when a docstyle definition setting has an
                               invalid format.
    :return:                   An iterator returning each DocumentationComment
                               found in the content.
    """
    docstyle_definition = DocstyleDefinition.load(language, docstyle)
    return extract_documentation_with_markers(content,
                                              docstyle_definition.markers)


1			import re
2
3			from coalib.bearlib.languages.documentation.DocstyleDefinition import (
4			DocstyleDefinition)
5			from coalib.bearlib.languages.documentation.DocumentationComment import (
6			DocumentationComment)
7			from coalib.results.TextRange import TextRange
8
9
10			def _extract_doccomment_simple(content, line, column, markers):
11			"""
12			Extract a documentation that starts at given beginning with simple layout.
13
14			The property of the simple layout is that there's no each-line marker. This
15			applies e.g. for python docstrings.
16
17			:param content: Presplitted lines of the source-code-string.
18			:param line: Line where the documentation comment starts (behind the
19			start marker). Zero-based.
20			:param column: Column where the documentation comment starts (behind the
21			start marker). Zero-based.
22			:param markers: The documentation identifying markers.
23			:return: If the comment matched layout a triple with end-of-comment
24			line, column and the extracted documentation. If not
25			matched, returns None.
26			"""
27			align_column = column - len(markers[0])
28
29			pos = content[line].find(markers[2], column)
30			if pos == -1:
31			doccomment = content[line][column:]
32			line += 1
33
34			while line < len(content):
35			pos = content[line].find(markers[2])
36			if pos == -1:
37			doccomment += content[line][align_column:]
38			else:
39			doccomment += content[line][align_column:pos]
40			return line, pos + len(markers[2]), doccomment
41
42			line += 1
43
44			return None
45			else:
46			return line, pos + len(markers[2]), content[line][column:pos]
47
48
49			def _extract_doccomment_continuous(content, line, column, markers):
50			"""
51			Extract a documentation that starts at given beginning with continuous
52			layout.
53
54			The property of the continuous layout is that the each-line-marker and the
55			end-marker do equal. Documentation is extracted until no further marker is
56			found. Applies e.g. for doxygen style python documentation:
57
58			```
59			## main
60			#
61			# detailed
62			```
63
64			:param content: Presplitted lines of the source-code-string.
65			:param line: Line where the documentation comment starts (behind the
66			start marker). Zero-based.
67			:param column: Column where the documentation comment starts (behind the
68			start marker). Zero-based.
69			:param markers: The documentation identifying markers.
70			:return: If the comment matched layout a triple with end-of-comment
71			line, column and the extracted documentation. If not
72			matched, returns None.
73			"""
74			marker_len = len(markers[1])
75
76			doccomment = content[line][column:]
77			line += 1
78			while line < len(content):
79			pos = content[line].find(markers[1])
80			if pos == -1:
81			return line, 0, doccomment
82			else:
83			doccomment += content[line][pos+marker_len:]
84
85			line += 1
86
87			if content[line - 1][-1] == "\n":
88			column = 0
89			else:
90			# This case can appear on end-of-document without a `\n`.
91			line -= 1
92			column = len(content[line])
93
94			return line, column, doccomment
95
96
97			def _extract_doccomment_standard(content, line, column, markers):
98			"""
99			Extract a documentation that starts at given beginning with standard
100			layout.
101
102			The standard layout applies e.g. for C doxygen-style documentation:
103
104			```
105			/**
106			* documentation
107			*/
108			```
109
110			:param content: Presplitted lines of the source-code-string.
111			:param line: Line where the documentation comment starts (behind the
112			start marker). Zero-based.
113			:param column: Column where the documentation comment starts (behind the
114			start marker). Zero-based.
115			:param markers: The documentation identifying markers.
116			:return: If the comment matched layout a triple with end-of-comment
117			line, column and the extracted documentation. If not
118			matched, returns None.
119			"""
120			pos = content[line].find(markers[2], column)
121			if pos != -1:
122			return line, pos + len(markers[2]), content[line][column:pos]
123			else:
124			doccomment = content[line][column:]
125			line += 1
126
127			while line < len(content):
128			pos = content[line].find(markers[2])
129			each_line_pos = content[line].find(markers[1])
130
131			if pos == -1:
132			if each_line_pos == -1:
133			# If the first text occurrence is not the each-line marker
134			# now we violate the doc-comment layout.
135			return None
136			doccomment += content[line][each_line_pos+len(markers[1]):]
137			else:
138			# If no each-line marker found or it's located past the end
139			# marker: extract no further and end the doc-comment.
140			if each_line_pos != -1 and each_line_pos + 1 < pos:
141			doccomment += (
142			content[line][each_line_pos+len(markers[1]):pos])
143
144			return line, pos + len(markers[2]), doccomment
145
146			line += 1
147
148			return None
149
150
151			def _extract_doccomment(content, line, column, markers):
152			"""
153			Delegates depending on the given markers to the right extraction method.
154
155			:param content: Presplitted lines of the source-code-string.
156			:param line: Line where the documentation comment starts (behind the
157			start marker). Zero-based.
158			:param column: Column where the documentation comment starts (behind the
159			start marker). Zero-based.
160			:param markers: The documentation identifying markers.
161			:return: If the comment matched layout a triple with end-of-comment
162			line, column and the extracted documentation. If not
163			matched, returns None.
164			"""
165			if markers[1] == "":
166			# Extract and align to start marker.
167			return _extract_doccomment_simple(content, line, column, markers)
168			elif markers[1] == markers[2]:
169			# Search for the each-line marker until it runs out.
170			return _extract_doccomment_continuous(content, line, column, markers)
171			else:
172			return _extract_doccomment_standard(content, line, column, markers)
173
174
175			def _compile_multi_match_regex(strings):
176			"""
177			Compiles a regex object that matches each of the given strings.
178
179			:param strings: The strings to match.
180			:return: A regex object.
181			"""
182			return re.compile("\|".join(re.escape(s) for s in strings))
183
184
185			def _extract_doccomment_from_line(content, line, column, regex, marker_dict):
186			begin_match = regex.search(content[line], column)
187			if begin_match:
188			column = begin_match.end()
189			for marker in marker_dict[begin_match.group()]:
190			doccomment = _extract_doccomment(content, line, column, marker)
191			if doccomment is not None:
192			end_line, end_column, documentation = doccomment
			0 ignored issues – show Bug Best Practice introduced 2015-12-16 20:03 UTC by Report Bug Copy Issue Report It seems like you are trying to unpack a non-sequence, which was defined at line 44. Loading history... Bug Best Practice introduced 2015-12-16 20:03 UTC by Report Bug Copy Issue Report It seems like you are trying to unpack a non-sequence, which was defined at line 135. Loading history... Bug Best Practice introduced 2015-12-16 20:03 UTC by Report Bug Copy Issue Report It seems like you are trying to unpack a non-sequence, which was defined at line 148. Loading history...
193
194			rng = TextRange.from_values(line + 1,
195			begin_match.start() + 1,
196			end_line + 1,
197			end_column + 1)
198			doc = DocumentationComment(documentation, marker, rng)
199
200			return end_line, end_column, doc
201
202			return line + 1, 0, None
203
204
205			def extract_documentation_with_markers(content, markers):
206			"""
207			Extracts all documentation texts inside the given source-code-string.
208
209			:param content: The source-code-string where to extract documentation from.
210			Needs to be a list or tuple where each string item is a
211			single line (including ending whitespaces like `\\n`).
212			:param markers: The list/tuple of marker-sets that identify a
213			documentation-comment. Low-index markers have higher
214			priority than high-index markers.
215			:return: An iterator returning each DocumentationComment found in
216			the content.
217			"""
218			# Prepare marker-tuple dict that maps a begin pattern to the corresponding
219			# marker_set(s). This makes it faster to retrieve a marker-set from a
220			# begin sequence we initially want to search for in source code. Then
221			# the possible found documentation match is processed further with the
222			# rest markers.
223			marker_dict = {}
224			for marker_set in markers:
225			if marker_set[0] not in marker_dict:
226			marker_dict[marker_set[0]] = [marker_set]
227			else:
228			marker_dict[marker_set[0]].append(marker_set)
229
230			# Using regexes to perform a variable match is faster than finding each
231			# substring with `str.find()` choosing the lowest match.
232			begin_regex = _compile_multi_match_regex(
233			marker_set[0] for marker_set in markers)
234
235			line = 0
236			column = 0
237			while line < len(content):
238			line, column, doc = _extract_doccomment_from_line(content,
239			line,
240			column,
241			begin_regex,
242			marker_dict)
243			if doc:
244			yield doc
245
246
247			def extract_documentation(content, language, docstyle):
248			"""
249			Extracts all documentation texts inside the given source-code-string using
250			the coala docstyle definition files.
251
252			The documentation texts are sorted by their order appearing in `content`.
253
254			For more information about how documentation comments are identified and
255			extracted, see DocstyleDefinition.doctypes enumeration.
256
257			:param content: The source-code-string where to extract
258			documentation from. Needs to be a list or tuple
259			where each string item is a single line
260			(including ending whitespaces like `\\n`).
261			:param language: The programming language used.
262			:param docstyle: The documentation style/tool used
263			(e.g. doxygen).
264			:raises FileNotFoundError: Raised when the docstyle definition file was not
265			found. This is a compatability exception from
266			`coalib.misc.Compatability` module.
267			:raises KeyError: Raised when the given language is not defined in
268			given docstyle.
269			:raises ValueError: Raised when a docstyle definition setting has an
270			invalid format.
271			:return: An iterator returning each DocumentationComment
272			found in the content.
273			"""
274			docstyle_definition = DocstyleDefinition.load(language, docstyle)
275			return extract_documentation_with_markers(content,
276			docstyle_definition.markers)
277

coala-analyzer / coala

Pull Request — master (#1139)

coalib.bearlib.languages.documentation.extract_documentation() B

Complexity

Size

Duplication

Duplication Side-by-Side

Filter issues like