Merge branch 'mborsetti-py310'

* mborsetti-py310:
  Added support Python 3.10, 3.11 and 3.12; removed older Python
  Support Python 3.10

.gitattributes

@@ -0,0 +1,4 @@
+*       text eol=lf
+*.bat   text eol=crlf
+*.png   -text
+*.pdf   -text

.travis.yml

@@ -0,0 +1,33 @@
+language: python
+cache: pip
+
+matrix:
+  include:
+    - python: 3.12
+      env: TOXENV=black
+    - python: 3.12
+      env: TOXENV=flake8
+    - python: 3.12
+      env: TOXENV=mypy
+    - python: 3.12
+      env: TOXENV=isort
+    - python: 3.8
+      env: TOXENV=py38
+    - python: 3.9
+      env: TOXENV=py39
+    - python: 3.10
+      env: TOXENV=py310
+    - python: 3.11
+      env: TOXENV=py310
+    - python: 3.12
+      env: TOXENV=py310
+    - python: pypy3
+      env: TOXENV=pypy3
+
+install:
+  - pip install tox
+script:
+  - tox
+after_success:
+  - pip install coveralls
+  - coveralls

html2text/__init__.py

@@ -9,8 +9,8 @@
 from typing import Dict, List, Optional, Tuple, Union
 
 from . import config
+from ._typing import OutCallback
 from .elements import AnchorElement, ListElement
-from .typing import OutCallback
 from .utils import (
     dumb_css_parser,
     element_style,
@@ -93,20 +93,20 @@ def __init__(
             self.out = out
 
         # empty list to store output characters before they are "joined"
-        self.outtextlist = []  # type: List[str]
+        self.outtextlist: List[str] = []
 
         self.quiet = 0
         self.p_p = 0  # number of newline character to print before next output
         self.outcount = 0
         self.start = True
         self.space = False
-        self.a = []  # type: List[AnchorElement]
-        self.astack = []  # type: List[Optional[Dict[str, Optional[str]]]]
-        self.maybe_automatic_link = None  # type: Optional[str]
+        self.a: List[AnchorElement] = []
+        self.astack: List[Optional[Dict[str, Optional[str]]]] = []
+        self.maybe_automatic_link: Optional[str] = None
         self.empty_link = False
         self.absolute_url_matcher = re.compile(r"^[a-zA-Z+]+://")
         self.acount = 0
-        self.list = []  # type: List[ListElement]
+        self.list: List[ListElement] = []
         self.blockquote = 0
         self.pre = False
         self.startpre = False
@@ -116,19 +116,17 @@ def __init__(
         self.lastWasNL = False
         self.lastWasList = False
         self.style = 0
-        self.style_def = {}  # type: Dict[str, Dict[str, str]]
-        self.tag_stack = (
-            []
-        )  # type: List[Tuple[str, Dict[str, Optional[str]], Dict[str, str]]]
+        self.style_def: Dict[str, Dict[str, str]] = {}
+        self.tag_stack: List[Tuple[str, Dict[str, Optional[str]], Dict[str, str]]] = []
         self.emphasis = 0
         self.drop_white_space = 0
         self.inheader = False
         # Current abbreviation definition
-        self.abbr_title = None  # type: Optional[str]
+        self.abbr_title: Optional[str] = None
         # Last inner HTML (for abbr being defined)
-        self.abbr_data = None  # type: Optional[str]
+        self.abbr_data: Optional[str] = None
         # Stack of abbreviations to write later
-        self.abbr_list = {}  # type: Dict[str, str]
+        self.abbr_list: Dict[str, str] = {}
         self.baseurl = baseurl
         self.stressed = False
         self.preceding_stressed = False
@@ -324,7 +322,7 @@ def handle_tag(
             # need the attributes of the parent nodes in order to get a
             # complete style description for the current element. we assume
             # that google docs export well formed html.
-            parent_style = {}  # type: Dict[str, str]
+            parent_style: Dict[str, str] = {}
             if start:
                 if self.tag_stack:
                     parent_style = self.tag_stack[-1][2]

setup.cfg

@@ -17,10 +17,11 @@ classifiers =
     Operating System :: OS Independent
     Programming Language :: Python
     Programming Language :: Python :: 3
-    Programming Language :: Python :: 3.7
     Programming Language :: Python :: 3.8
     Programming Language :: Python :: 3.9
     Programming Language :: Python :: 3.10
+    Programming Language :: Python :: 3.11
+    Programming Language :: Python :: 3.12
     Programming Language :: Python :: 3 :: Only
     Programming Language :: Python :: Implementation :: CPython
     Programming Language :: Python :: Implementation :: PyPy
@@ -29,7 +30,7 @@ platform = OS Independent
 [options]
 zip_safe = False
 packages = html2text
-python_requires = >=3.7
+python_requires = >=3.8
 
 [options.entry_points]
 console_scripts =
@@ -47,4 +48,4 @@ combine_as_imports = True
 profile = black
 
 [mypy]
-python_version = 3.7
+python_version = 3.8

test/abbr_tag.md

@@ -1,4 +1,3 @@
 TLA xyz
 
   *[TLA]: Three Letter Acronym
-

test/anchors.md

@@ -5,4 +5,3 @@ Additional hyperlink tests!
 [**Bold Link**](http://some.link)
 [`filename.py`](http://some.link/filename.py) [The source code is called
 `magic.py`](http://some.link/magicsources.py)
-

test/apos_element.md

@@ -1,2 +1 @@
 '
-

test/blockquote_example.md

@@ -1,4 +1,3 @@
 > "The time has come", the Walrus said, "To talk of many things: Of shoes -
 > and ships - and sealing wax - Of cabbages - and kings- And why the sea is
 > boiling hot - And whether pigs have wings."
-

test/bold_long_line.md

@@ -1,3 +1,2 @@
 **text** and a very long long long long long long long long long long long
 long long long long long long long long long line
-

test/br_inside_a.md

@@ -0,0 +1 @@
+[This is a test](https://example.org/)

test/break_preserved_in_blockquote.md

@@ -2,4 +2,3 @@ a
 
 > b  
 > c
-

test/css_import_no_semicolon.md

@@ -2,4 +2,3 @@
 
 Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod
 tempor incididunt ut labore et dolore magna aliqua.
-

test/decript_tage.md

@@ -1,2 +1 @@
 ~~something~~ ~~something~~ ~~something~~
-

test/default_image_alt.md

@@ -1,2 +1 @@
 [![Image](images/google.png)](http://google.com)
-

test/doc_with_table.md

@@ -10,4 +10,3 @@ Header 1 | Header 2 | Header 3
 ---|---|---  
 Content 1 | Content 2 | ![200](http://lorempixel.com/200/200) Image!  
 Content 1 | Content 2 | ![200](http://lorempixel.com/200/200) Image!
-

test/doc_with_table_bypass.md

@@ -39,4 +39,3 @@ Content 2
 ![200](http://lorempixel.com/200/200) Image!
 
 </td> </tr> </table>
-

test/emdash-para.md

@@ -12,4 +12,3 @@ Duis brisket drumstick, dolor fatback filet mignon meatloaf laboris tri-tip
 speck chuck ball tip voluptate ullamco laborum.
 
 \--
-

test/emphasis_preserved_whitespace.md

@@ -35,4 +35,3 @@ _emphasis_!
 ~ ~~strike~~
 
 _em1_ _em2_
-

test/emphasis_whitespace.md

@@ -29,4 +29,3 @@ _~~si~~_
 **_~~bis~~_**
 
 .**_~~bis~~_**
-

test/empty-link.md

@@ -3,4 +3,3 @@
 This test checks whether empty hyperlinks still appear in the markdown result.
 
 [](http://some.link) [](http://some.link)
-

test/empty-title-tag.md

@@ -1,2 +1 @@
 [This is an A tag with an empty title property](test.html)
-

test/flip_emphasis.md

@@ -1,2 +1 @@
 *Something* __else__
-

test/header_tags.md

@@ -17,4 +17,3 @@
 ######### H9
 
 H10 H11 H12 NO number
-

test/horizontal_rule.md

@@ -1,2 +1 @@
 * * *
-

test/html-escaping.md

@@ -5,4 +5,3 @@ Escaped HTML like <div> or & should NOT remain escaped on output
     ...even when that escaped HTML is in a <pre> tag
 
 `...or a <code> tag`
-

test/html_entities_out_of_text.md

@@ -1,2 +1 @@
 [allas: Country Manager](http://thth)
-

test/images_as_html.md

@@ -4,4 +4,3 @@ src='image_with_width.jpg' height='300' alt='An image with a height attr' />
 <img src='image_with_width_and_height.jpg' width='300' height='300' alt='An
 image with width and height' /> <img src='image_with_width_and_height.jpg'
 width='300' height='300' /> <img src='image_with_width_and_height.jpg' />
-

test/images_with_size.md

@@ -4,4 +4,3 @@ src='image_with_width.jpg' width='300' alt='An image with a width attr' />
 /> <img src='image_with_width_and_height.jpg' width='300' height='300' alt='An
 image with width and height' /> <img src='image_with_width_and_height.jpg'
 width='300' height='300' /> ![](image_with_width_and_height.jpg)
-

test/img-tag-with-link.md

@@ -7,4 +7,3 @@ This test checks images with associated links.
 [![{banana}](http://placehold.it/350x150#{banana})](http://some.link)
 [![\(\[{}\]\)](http://placehold.it/350x150#\(\[{}\]\))](http://some.link)
 [![](http://placehold.it/350x150#\(\[{}\]\))](http://some.link)
-

test/invalid_start.md

@@ -1,3 +1,2 @@
   1. The ol has an invalid start 
   2. This should just be ignored 
-

test/invalid_unicode.md

@@ -1,2 +1 @@
 Br
-

test/kbd_tag.md

@@ -1,2 +1 @@
 Press `[CTRL]+c` to copy.
-

test/link_titles.md

@@ -1,3 +1,2 @@
 [ first example](http://example.com "MyTitle")  
 [ second example](http://example.com)
-

test/list_tags_example.md

@@ -35,4 +35,3 @@ Drums
 
   * something else here
   * some item
-

test/long_lines.md

@@ -11,4 +11,3 @@ asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd
 asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd
 asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd asd
 asd asd asd asd asd
-

test/lrm_after_b.md

@@ -1,2 +1 @@
 **b**
-

test/lrm_after_i.md

@@ -1,2 +1 @@
 _Foo_
-

test/lrm_inside_i.md

@@ -1,2 +1 @@
 _Foo bar_
-

test/mark_code.md

@@ -10,4 +10,3 @@ Normal text with 'pre' code block.
 [/code]
 
 Normal text continues.
-

test/mixed_nested_lists.md

@@ -9,4 +9,3 @@
     1. ordered
     2. ...
   * end
-

test/nbsp.md

@@ -11,4 +11,3 @@ consequat.
 Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore
 eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt
 in culpa qui officia deserunt mollit anim id est laborum.
-

test/nbsp_unicode.md

@@ -11,4 +11,3 @@ consequat.
 Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore
 eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt
 in culpa qui officia deserunt mollit anim id est laborum.
-

test/no_inline_links_example.md

@@ -6,4 +6,3 @@
    [2]: http://example.com (Example title)
 
    [3]: http://example.com (abc)
-

test/no_inline_links_images_to_alt.md

@@ -5,4 +5,3 @@
    [1]: http://example.com/img.png
 
    [2]: http://example.com
-

test/no_mailto_links.md

@@ -1,2 +1 @@
 Send an email to [email protected].
-

test/no_wrap_links.md

@@ -1,2 +1 @@
 And [here](http://bugs.debian.org/cgi-bin/pkgreport.cgi?tag=multiarch;[email protected]) is a long link I had at hand.
-

test/no_wrap_links_no_inline_links.md

@@ -1,2 +1 @@
 And [here](http://bugs.debian.org/cgi-bin/pkgreport.cgi?tag=multiarch;[email protected]) is a long link I had at hand. 
-

test/normal.md

@@ -51,4 +51,3 @@ not a hr
 \- - -
 
 c:\tmp, \\\server\path, \\_/, foo\bar, #\\#, \\\\#
-

test/normal_escape_snob.md

@@ -52,4 +52,3 @@ not a hr
 \- - -
 
 c:\tmp, \\\server\path, \\\_/, foo\bar, \#\\\#, \\\\\#
-

test/pad_table.md

@@ -38,5 +38,3 @@ something else entirely
 | A       | B+C   |
 | A+B     | C     |
 | A+B+C   |
-
-

test/pad_table_empty.md

@@ -1,3 +1 @@
 |  |
-
-

test/pad_table_no_closed_tr.md

@@ -1,3 +1 @@
 | **xxx** | **xxx** | xxxx | **xxxx** | **xxxx** | xxxx |
-
-

test/pre.md

@@ -5,4 +5,3 @@
 
 Ensure that HTML that starts with a crowded `<pre>` is converted to reasonable
 Markdown.
-

test/preformatted_in_list.md

@@ -4,4 +4,3 @@
 
   * ?
   * Profit!
-

test/protect_links.md

@@ -1,3 +1,2 @@
 [foo](<http://im-a-very-very-very-very-very-very-very-very-very-very-
 long/link.html>)
-

test/q_tag.md

@@ -1,2 +1 @@
 "If this is a test," he said, "then it should pass".
-

test/rlm_inside_strong.md

@@ -1,2 +1 @@
 **Foo bar**
-

test/stressed_with_html_entities.md

@@ -1,2 +1 @@
 **hello** world ><
-

test/table_ignore.md

@@ -19,4 +19,3 @@ Header 1 Header 2 Header 3
 Content 1 Content 2 ![200](http://lorempixel.com/200/200) Image!  
 Content 1 Content 2 longer ![200](http://lorempixel.com/200/200) Image!  
 something else entirely
-