Coverage for sources/mimeogram/acquirers.py: 92%
133 statements
« prev ^ index » next coverage.py v7.6.12, created at 2025-03-02 23:41 +0000
1# vim: set filetype=python fileencoding=utf-8:
2# -*- coding: utf-8 -*-
4#============================================================================#
5# #
6# Licensed under the Apache License, Version 2.0 (the "License"); #
7# you may not use this file except in compliance with the License. #
8# You may obtain a copy of the License at #
9# #
10# http://www.apache.org/licenses/LICENSE-2.0 #
11# #
12# Unless required by applicable law or agreed to in writing, software #
13# distributed under the License is distributed on an "AS IS" BASIS, #
14# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. #
15# See the License for the specific language governing permissions and #
16# limitations under the License. #
17# #
18#============================================================================#
21''' Content acquisition from various sources. '''
24from __future__ import annotations
26import aiofiles as _aiofiles
27import httpx as _httpx
29from . import __
30from . import exceptions as _exceptions
31from . import parts as _parts
34_scribe = __.produce_scribe( __name__ )
37async def acquire( # pylint: disable=too-many-locals
38 auxdata: __.Globals, sources: __.cabc.Sequence[ str | __.Path ]
39) -> __.cabc.Sequence[ _parts.Part ]:
40 ''' Acquires content from multiple sources. '''
41 from urllib.parse import urlparse
42 options = auxdata.configuration.get( 'acquire-parts', { } )
43 strict = options.get( 'fail-on-invalid', False )
44 recursive = options.get( 'recurse-directories', False )
45 tasks: list[ __.cabc.Coroutine[ None, None, _parts.Part ] ] = [ ]
46 for source in sources:
47 path = __.Path( source )
48 url_parts = (
49 urlparse( source ) if isinstance( source, str )
50 else urlparse( str( source ) ) )
51 scheme = 'file' if path.drive else url_parts.scheme
52 match scheme:
53 case '' | 'file':
54 tasks.extend( _produce_fs_tasks( source, recursive ) )
55 case 'http' | 'https':
56 tasks.append( _produce_http_task( str( source ) ) )
57 case _:
58 raise _exceptions.UrlSchemeNoSupport( str( source ) )
59 if strict: return await __.gather_async( *tasks )
60 results = await __.gather_async( *tasks, return_exceptions = True )
61 # TODO: Factor into '__.generics.extract_results_filter_errors'.
62 values: list[ _parts.Part ] = [ ]
63 for result in results:
64 if result.is_error( ):
65 _scribe.warning( str( result.error ) )
66 continue
67 values.append( result.extract( ) )
68 return tuple( values )
71async def _acquire_from_file( location: __.Path ) -> _parts.Part:
72 ''' Acquires content from text file. '''
73 from .exceptions import ContentAcquireFailure, ContentDecodeFailure
74 try:
75 async with _aiofiles.open( location, 'rb' ) as f:
76 content_bytes = await f.read( )
77 except Exception as exc: raise ContentAcquireFailure( location ) from exc
78 mimetype, charset = _detect_mimetype_and_charset( content_bytes, location )
79 if charset is None: raise ContentDecodeFailure( location, '???' ) 79 ↛ exitline 79 didn't except from function '_acquire_from_file' because the raise on line 79 wasn't executed
80 linesep = _parts.LineSeparators.detect_bytes( content_bytes )
81 # TODO? Separate error for newline issues.
82 if linesep is None: raise ContentDecodeFailure( location, charset ) 82 ↛ exitline 82 didn't except from function '_acquire_from_file' because the raise on line 82 wasn't executed
83 try: content = content_bytes.decode( charset )
84 except Exception as exc:
85 raise ContentDecodeFailure( location, charset ) from exc
86 _scribe.debug( f"Read file: {location}" )
87 return _parts.Part(
88 location = str( location ),
89 mimetype = mimetype,
90 charset = charset,
91 linesep = linesep,
92 content = linesep.normalize( content ) )
95async def _acquire_via_http( # pylint: disable=too-many-locals
96 client: _httpx.AsyncClient, url: str
97) -> _parts.Part:
98 ''' Acquires content via HTTP/HTTPS. '''
99 from .exceptions import ContentAcquireFailure, ContentDecodeFailure
100 try:
101 response = await client.get( url )
102 response.raise_for_status( )
103 except Exception as exc: raise ContentAcquireFailure( url ) from exc
104 mimetype = (
105 response.headers.get( 'content-type', 'application/octet-stream' )
106 .split( ';' )[ 0 ].strip( ) )
107 content_bytes = response.content
108 charset = response.encoding or _detect_charset( content_bytes )
109 if charset is None: raise ContentDecodeFailure( url, '???' ) 109 ↛ exitline 109 didn't except from function '_acquire_via_http' because the raise on line 109 wasn't executed
110 if not _is_textual_mimetype( mimetype ):
111 mimetype, _ = (
112 _detect_mimetype_and_charset(
113 content_bytes, url, charset = charset ) )
114 linesep = _parts.LineSeparators.detect_bytes( content_bytes )
115 # TODO? Separate error for newline issues.
116 if linesep is None: raise ContentDecodeFailure( url, charset ) 116 ↛ exitline 116 didn't except from function '_acquire_via_http' because the raise on line 116 wasn't executed
117 try: content = content_bytes.decode( charset )
118 except Exception as exc:
119 raise ContentDecodeFailure( url, charset ) from exc
120 _scribe.debug( f"Fetched URL: {url}" )
121 return _parts.Part(
122 location = url,
123 mimetype = mimetype,
124 charset = charset,
125 linesep = linesep,
126 content = linesep.normalize( content ) )
129# VCS directories to skip during traversal
130_VCS_DIRS = frozenset( ( '.git', '.svn', '.hg', '.bzr' ) )
131def _collect_directory_files(
132 directory: __.Path, recursive: bool
133) -> list[ __.Path ]:
134 ''' Collects and filters files from directory hierarchy. '''
135 import gitignorefile
136 cache = gitignorefile.Cache( )
137 paths: list[ __.Path ] = [ ]
138 _scribe.debug( f"Collecting files in directory: {directory}" )
139 for entry in directory.iterdir( ):
140 if entry.is_dir( ) and entry.name in _VCS_DIRS:
141 _scribe.debug( f"Ignoring VCS directory: {entry}" )
142 continue
143 if cache( str( entry ) ):
144 _scribe.debug( f"Ignoring path (matched by .gitignore): {entry}" )
145 continue
146 if entry.is_dir( ) and recursive:
147 paths.extend( _collect_directory_files( entry, recursive ) )
148 elif entry.is_file( ): paths.append( entry )
149 return paths
152def _detect_charset( content: bytes ) -> str | None:
153 # TODO: Pyright bug: `None is charset` != `charset is None`
154 from chardet import detect
155 charset = detect( content )[ 'encoding' ]
156 if charset is None: return charset 156 ↛ exitline 156 didn't return from function '_detect_charset' because the return on line 156 wasn't executed
157 if charset.startswith( 'utf' ): return charset
158 match charset:
159 case 'ascii': return 'utf-8' # Assume superset.
160 case _: pass
161 # Shake out false positives, like 'MacRoman'.
162 try: content.decode( 'utf-8' )
163 except UnicodeDecodeError: return charset
164 return 'utf-8'
167def _detect_mimetype( content: bytes, location: str | __.Path ) -> str | None:
168 from mimetypes import guess_type
169 from puremagic import PureError, from_string # pyright: ignore
170 try: return from_string( content, mime = True )
171 except ( PureError, ValueError ):
172 return guess_type( str( location ) )[ 0 ]
175def _detect_mimetype_and_charset(
176 content: bytes,
177 location: str | __.Path, *,
178 mimetype: __.Absential[ str ] = __.absent,
179 charset: __.Absential[ str ] = __.absent,
180) -> tuple[ str, str | None ]:
181 from .exceptions import TextualMimetypeInvalidity
182 if __.is_absent( mimetype ): 182 ↛ 184line 182 didn't jump to line 184 because the condition on line 182 was always true
183 mimetype_ = _detect_mimetype( content, location )
184 else: mimetype_ = mimetype
185 if __.is_absent( charset ):
186 charset_ = _detect_charset( content )
187 else: charset_ = charset
188 if not mimetype_:
189 if charset_: mimetype_ = 'text/plain' # pylint: disable=redefined-variable-type 189 ↛ 190line 189 didn't jump to line 190 because the condition on line 189 was always true
190 else: mimetype_ = 'application/octet-stream'
191 if not _is_textual_mimetype( mimetype_ ):
192 raise TextualMimetypeInvalidity( location, mimetype_ )
193 return mimetype_, charset_
196# MIME types that are considered textual beyond those starting with 'text/'
197_TEXTUAL_MIME_TYPES = frozenset( (
198 'application/json',
199 'application/xml',
200 'application/xhtml+xml',
201 'application/javascript',
202 'image/svg+xml',
203) )
204def _is_textual_mimetype( mimetype: str ) -> bool:
205 ''' Checks if MIME type represents textual content. '''
206 _scribe.debug( f"MIME type: {mimetype}" )
207 if mimetype.startswith( ( 'text/', 'application/x-', 'text/x-' ) ):
208 return True
209 return mimetype in _TEXTUAL_MIME_TYPES
212def _produce_fs_tasks(
213 location: str | __.Path, recursive: bool = False
214) -> tuple[ __.cabc.Coroutine[ None, None, _parts.Part ], ...]:
215 location_ = __.Path( location )
216 if location_.is_file( ) or location_.is_symlink( ):
217 return ( _acquire_from_file( location_ ), )
218 if location_.is_dir( ):
219 files = _collect_directory_files( location_, recursive )
220 return tuple( _acquire_from_file( f ) for f in files )
221 raise _exceptions.ContentAcquireFailure( location )
224def _produce_http_task(
225 url: str
226) -> __.cabc.Coroutine[ None, None, _parts.Part ]:
227 # TODO: URL object rather than string.
228 # TODO: Reuse clients for common hosts.
230 async def _execute_session( ) -> _parts.Part:
231 async with _httpx.AsyncClient( # nosec B113
232 follow_redirects = True
233 ) as client: return await _acquire_via_http( client, url )
235 return _execute_session( )