Add support for HTML handling

daniel-jones-dev · daniel-jones-dev · commit 9b88f6a0fe89 · 2022-02-15T22:17:15.000+01:00
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,6 +7,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 ### Added
+* Add support for HTML tag handling in `translate_text()`.
 ### Changed
 ### Deprecated
 ### Removed
diff --git a/deepl/__main__.py b/deepl/__main__.py
@@ -289,7 +289,7 @@ def add_common_arguments(subparser: argparse.ArgumentParser):
     tag_handling_group.add_argument(
         "--tag-handling",
         type=str,
-        choices=["xml"],
+        choices=["xml", "html"],
         default=None,
         help="activate processing of formatting tags, for example 'xml'",
     )
diff --git a/deepl/translator.py b/deepl/translator.py
@@ -639,17 +639,19 @@ def translate_text(
         :param glossary: (Optional) glossary or glossary ID to use for
             translation. Must match specified source_lang and target_lang.
         :param tag_handling: (Optional) Type of tags to parse before
-            translation, only "xml" is currently available.
+            translation, only "xml" and "html" are currently available.
         :param outline_detection: (Optional) Set to False to disable automatic
             tag detection.
-        :param non_splitting_tags: (Optional) Tags that should not split a
+        :param non_splitting_tags: (Optional) XML tags that should not split a
             sentence.
-        :type non_splitting_tags: List of tags or comma-separated-list of tags.
-        :param splitting_tags: (Optional) Tags that should split a sentence.
-        :type splitting_tags: List of tags or comma-separated-list of tags.
-        :param ignore_tags: (Optional) Tags containing text that should not be
-            translated.
-        :type ignore_tags: List of tags or comma-separated-list of tags.
+        :type non_splitting_tags: List of XML tags or comma-separated-list of
+            tags.
+        :param splitting_tags: (Optional) XML tags that should split a
+            sentence.
+        :type splitting_tags: List of XML tags or comma-separated-list of tags.
+        :param ignore_tags: (Optional) XML tags containing text that should not
+            be translated.
+        :type ignore_tags: List of XML tags or comma-separated-list of tags.
         :return: List of TextResult objects containing results, unless input
             text was one string, then a single TextResult object is returned.
         """
diff --git a/tests/test_cli.py b/tests/test_cli.py
@@ -147,6 +147,14 @@ def test_text_tags(runner):
     ), f"output:\n{result.output}"
 
 
+def test_text_html_tag_handling(runner):
+    result = runner.invoke(
+        main_function,
+        '-vv text --to DE --tag-handling html "<html><p>Test</p></html>"',
+    )
+    assert result.exit_code == 0, f"exit: {result.exit_code}\n {result.output}"
+
+
 def test_document(runner, tmpdir):
     tmpdir = pathlib.Path(tmpdir)
     output_dir = tmpdir / "output"
diff --git a/tests/test_translate_text.py b/tests/test_translate_text.py
@@ -259,6 +259,25 @@ def test_tag_handling_specify_tags(translator):
     assert re.compile("<title>.*Der Titel.*</title>").search(result.text)
 
 
+@needs_real_server
+def test_tag_handling_html(translator):
+    text = """
+<!DOCTYPE html>
+<html>
+   <body>
+       <h1>My First Heading</h1>
+       <p translate="no">My first paragraph.</p>
+   </body>
+</html>
+"""
+
+    result = translator.translate_text(
+        text, target_lang="DE", tag_handling="html"
+    )
+    assert "<h1>Meine erste Überschrift</h1>" in result.text
+    assert '<p translate="no">My first paragraph.</p>' in result.text
+
+
 def test_invalid_url(server):
     translator = deepl.Translator(
         server.auth_key, server_url="https://example.com"

Original file line number	Diff line number	Diff line change
`@@ -289,7 +289,7 @@ def add_common_arguments(subparser: argparse.ArgumentParser):`
`289`	`289`	`tag_handling_group.add_argument(`
`290`	`290`	`"--tag-handling",`
`291`	`291`	`type=str,`
`292`		`- choices=["xml"],`
	`292`	`+ choices=["xml", "html"],`
`293`	`293`	`default=None,`
`294`	`294`	`help="activate processing of formatting tags, for example 'xml'",`
`295`	`295`	`)`