Coverage for sources/librovore/structures/sphinx/detection.py: 25%

1# vim: set filetype=python fileencoding=utf-8: 

2# -*- coding: utf-8 -*- 

3 

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#============================================================================# 

19 

20 

21''' Sphinx detection and metadata extraction. ''' 

22 

23 

24from urllib.parse import ParseResult as _Url 

25 

26from . import __ 

27from . import extraction as _extraction 

28from . import urls as _urls 

29 

30 

31_scribe = __.acquire_scribe( __name__ ) 

32 

33 

34class SphinxDetection( __.StructureDetection ): 

35 ''' Detection result for Sphinx documentation sources. ''' 

36 

37 source: str 

38 has_searchindex: bool = False 

39 normalized_source: str = '' 

40 theme: __.typx.Optional[ str ] = None 

41 

42 @classmethod 

43 async def from_source( 

44 selfclass, 

45 auxdata: __.ApplicationGlobals, 

46 processor: __.Processor, 

47 source: str, 

48 ) -> __.typx.Self: 

49 ''' Constructs detection from source location. ''' 

50 detection = await processor.detect( auxdata, source ) 

51 return __.typx.cast( __.typx.Self, detection ) 

52 

53 async def extract_contents( 

54 self, 

55 auxdata: __.ApplicationGlobals, 

56 source: str, 

57 objects: __.cabc.Sequence[ __.InventoryObject ], /, 

58 ) -> tuple[ __.ContentDocument, ... ]: 

59 ''' Extracts documentation content for specified objects. ''' 

60 theme = self.theme if self.theme is not None else __.absent 

61 documents = await _extraction.extract_contents( 

62 auxdata, source, objects, theme = theme ) 

63 return tuple( documents ) 

64 

65 

66 

67async def check_searchindex( 

68 auxdata: __.ApplicationGlobals, source: _Url 

69) -> bool: 

70 ''' Checks if searchindex.js exists (indicates full Sphinx site). ''' 

71 url = _urls.derive_searchindex_url( source ) 

72 return await __.probe_url( auxdata.probe_cache, url ) 

73 

74 

75async def detect_theme( 

76 auxdata: __.ApplicationGlobals, source: _Url 

77) -> dict[ str, __.typx.Any ]: 

78 ''' Detects Sphinx theme and other metadata. ''' 

79 theme_metadata: dict[ str, __.typx.Any ] = { } 

80 html_url = _urls.derive_html_url( source ) 

81 try: 

82 # TODO: Use probe_url instead of `try`. 

83 html_content = await __.retrieve_url_as_text( 

84 auxdata.content_cache, 

85 html_url, duration_max = 10.0 ) 

86 except __.DocumentationInaccessibility: pass 

87 else: 

88 html_content_lower = html_content.lower( ) 

89 if ( 'furo' in html_content_lower 

90 or 'css/furo.css' in html_content_lower 

91 ): theme_metadata[ 'theme' ] = 'furo' 

92 elif ( 'sphinx_rtd_theme' in html_content_lower 

93 or 'css/theme.css' in html_content_lower 

94 ): theme_metadata[ 'theme' ] = 'sphinx_rtd_theme' 

95 elif ( 'alabaster' in html_content_lower 

96 or 'css/alabaster.css' in html_content_lower 

97 ): theme_metadata[ 'theme' ] = 'alabaster' 

98 elif ( 'pydoctheme.css' in html_content_lower 

99 or 'classic.css' in html_content_lower 

100 ): theme_metadata[ 'theme' ] = 'pydoctheme' 

101 elif 'flask.css' in html_content_lower: 

102 theme_metadata[ 'theme' ] = 'flask' 

103 elif 'css/nature.css' in html_content_lower: 

104 theme_metadata[ 'theme' ] = 'nature' 

105 elif 'css/default.css' in html_content_lower: 

106 theme_metadata[ 'theme' ] = 'classic' 

107 elif 'sphinx_book_theme' in html_content_lower: 

108 theme_metadata[ 'theme' ] = 'sphinx_book_theme' 

109 elif 'pydata_sphinx_theme' in html_content_lower: 

110 theme_metadata[ 'theme' ] = 'pydata_sphinx_theme' 

111 # If no theme detected, don't set theme key (returns None) 

112 return theme_metadata