1 files changed, 235 insertions, 0 deletions
diff --git a/tests/test_builders/test_build_gettext.py b/tests/test_builders/test_build_gettext.py
new file mode 100644
index 0000000..ddc6d30
--- /dev/null
+++ b/tests/test_builders/test_build_gettext.py
@@ -0,0 +1,235 @@
+"""Test the build process with gettext builder with the test root."""
+
+import gettext
+import os
+import re
+import subprocess
+import sys
+from subprocess import CalledProcessError
+
+import pytest
+
+from sphinx.builders.gettext import Catalog, MsgOrigin
+
+if sys.version_info[:2] >= (3, 11):
+    from contextlib import chdir
+else:
+    from sphinx.util.osutil import _chdir as chdir
+
+_MSGID_PATTERN = re.compile(r'msgid "(.*)"')
+
+
+def msgid_getter(msgid):
+    if m := _MSGID_PATTERN.search(msgid):
+        return m[1]
+    return None
+
+
+def test_Catalog_duplicated_message():
+    catalog = Catalog()
+    catalog.add('hello', MsgOrigin('/path/to/filename', 1))
+    catalog.add('hello', MsgOrigin('/path/to/filename', 1))
+    catalog.add('hello', MsgOrigin('/path/to/filename', 2))
+    catalog.add('hello', MsgOrigin('/path/to/yetanother', 1))
+    catalog.add('world', MsgOrigin('/path/to/filename', 1))
+
+    assert len(list(catalog)) == 2
+
+    msg1, msg2 = list(catalog)
+    assert msg1.text == 'hello'
+    assert msg1.locations == [('/path/to/filename', 1),
+                              ('/path/to/filename', 2),
+                              ('/path/to/yetanother', 1)]
+    assert msg2.text == 'world'
+    assert msg2.locations == [('/path/to/filename', 1)]
+
+
+@pytest.mark.sphinx('gettext', srcdir='root-gettext')
+def test_build_gettext(app):
+    # Generic build; should fail only when the builder is horribly broken.
+    app.build(force_all=True)
+
+    # Do messages end up in the correct location?
+    # top-level documents end up in a message catalog
+    assert (app.outdir / 'extapi.pot').is_file()
+    # directory items are grouped into sections
+    assert (app.outdir / 'subdir.pot').is_file()
+
+    # regression test for issue #960
+    catalog = (app.outdir / 'markup.pot').read_text(encoding='utf8')
+    assert 'msgid "something, something else, something more"' in catalog
+
+
+@pytest.mark.sphinx('gettext', srcdir='root-gettext')
+def test_msgfmt(app):
+    app.build(force_all=True)
+
+    (app.outdir / 'en' / 'LC_MESSAGES').mkdir(parents=True, exist_ok=True)
+    with chdir(app.outdir):
+        try:
+            args = ['msginit', '--no-translator', '-i', 'markup.pot', '--locale', 'en_US']
+            subprocess.run(args, capture_output=True, check=True)
+        except OSError:
+            pytest.skip()  # most likely msginit was not found
+        except CalledProcessError as exc:
+            print(exc.stdout)
+            print(exc.stderr)
+            msg = f'msginit exited with return code {exc.returncode}'
+            raise AssertionError(msg) from exc
+
+        assert (app.outdir / 'en_US.po').is_file(), 'msginit failed'
+        try:
+            args = ['msgfmt', 'en_US.po',
+                    '-o', os.path.join('en', 'LC_MESSAGES', 'test_root.mo')]
+            subprocess.run(args, capture_output=True, check=True)
+        except OSError:
+            pytest.skip()  # most likely msgfmt was not found
+        except CalledProcessError as exc:
+            print(exc.stdout)
+            print(exc.stderr)
+            msg = f'msgfmt exited with return code {exc.returncode}'
+            raise AssertionError(msg) from exc
+
+        mo = app.outdir / 'en' / 'LC_MESSAGES' / 'test_root.mo'
+        assert mo.is_file(), 'msgfmt failed'
+
+    _ = gettext.translation('test_root', app.outdir, languages=['en']).gettext
+    assert _("Testing various markup") == "Testing various markup"
+
+
+@pytest.mark.sphinx(
+    'gettext', testroot='intl', srcdir='gettext',
+    confoverrides={'gettext_compact': False})
+def test_gettext_index_entries(app):
+    # regression test for #976
+    app.build(filenames=[app.srcdir / 'index_entries.txt'])
+
+    pot = (app.outdir / 'index_entries.pot').read_text(encoding='utf8')
+    msg_ids = list(filter(None, map(msgid_getter, pot.splitlines())))
+
+    assert msg_ids == [
+        "i18n with index entries",
+        "index target section",
+        "this is :index:`Newsletter` target paragraph.",
+        "various index entries",
+        "That's all.",
+        "Mailing List",
+        "Newsletter",
+        "Recipients List",
+        "First",
+        "Second",
+        "Third",
+        "Entry",
+        "See",
+    ]
+
+
+@pytest.mark.sphinx(
+    'gettext', testroot='intl', srcdir='gettext',
+    confoverrides={'gettext_compact': False,
+                   'gettext_additional_targets': []})
+def test_gettext_disable_index_entries(app):
+    # regression test for #976
+    app.env._pickled_doctree_cache.clear()  # clear cache
+    app.build(filenames=[app.srcdir / 'index_entries.txt'])
+
+    pot = (app.outdir / 'index_entries.pot').read_text(encoding='utf8')
+    msg_ids = list(filter(None, map(msgid_getter, pot.splitlines())))
+
+    assert msg_ids == [
+        "i18n with index entries",
+        "index target section",
+        "this is :index:`Newsletter` target paragraph.",
+        "various index entries",
+        "That's all.",
+    ]
+
+
+@pytest.mark.sphinx('gettext', testroot='intl', srcdir='gettext')
+def test_gettext_template(app):
+    app.build(force_all=True)
+
+    assert (app.outdir / 'sphinx.pot').is_file()
+
+    result = (app.outdir / 'sphinx.pot').read_text(encoding='utf8')
+    assert "Welcome" in result
+    assert "Sphinx %(version)s" in result
+
+
+@pytest.mark.sphinx('gettext', testroot='gettext-template')
+def test_gettext_template_msgid_order_in_sphinxpot(app):
+    app.build(force_all=True)
+    assert (app.outdir / 'sphinx.pot').is_file()
+
+    result = (app.outdir / 'sphinx.pot').read_text(encoding='utf8')
+    assert re.search(
+        ('msgid "Template 1".*'
+         'msgid "This is Template 1\\.".*'
+         'msgid "Template 2".*'
+         'msgid "This is Template 2\\.".*'),
+        result,
+        flags=re.DOTALL)
+
+
+@pytest.mark.sphinx(
+    'gettext', srcdir='root-gettext',
+    confoverrides={'gettext_compact': 'documentation'})
+def test_build_single_pot(app):
+    app.build(force_all=True)
+
+    assert (app.outdir / 'documentation.pot').is_file()
+
+    result = (app.outdir / 'documentation.pot').read_text(encoding='utf8')
+    assert re.search(
+        ('msgid "Todo".*'
+         'msgid "Like footnotes.".*'
+         'msgid "The minute.".*'
+         'msgid "Generated section".*'),
+        result,
+        flags=re.DOTALL)
+
+
+@pytest.mark.sphinx(
+    'gettext',
+    testroot='intl_substitution_definitions',
+    srcdir='gettext-subst',
+    confoverrides={'gettext_compact': False,
+                   'gettext_additional_targets': ['image']})
+def test_gettext_prolog_epilog_substitution(app):
+    app.build(force_all=True)
+
+    assert (app.outdir / 'prolog_epilog_substitution.pot').is_file()
+    pot = (app.outdir / 'prolog_epilog_substitution.pot').read_text(encoding='utf8')
+    msg_ids = list(filter(None, map(msgid_getter, pot.splitlines())))
+
+    assert msg_ids == [
+        "i18n with prologue and epilogue substitutions",
+        "This is content that contains |subst_prolog_1|.",
+        "Substituted image |subst_prolog_2| here.",
+        "subst_prolog_2",
+        ".. image:: /img.png",
+        "This is content that contains |subst_epilog_1|.",
+        "Substituted image |subst_epilog_2| here.",
+        "subst_epilog_2",
+        ".. image:: /i18n.png",
+    ]
+
+
+@pytest.mark.sphinx(
+    'gettext',
+    testroot='intl_substitution_definitions',
+    srcdir='gettext-subst',
+    confoverrides={'gettext_compact': False,
+                   'gettext_additional_targets': ['image']})
+def test_gettext_prolog_epilog_substitution_excluded(app):
+    # regression test for #9428
+    app.build(force_all=True)
+
+    assert (app.outdir / 'prolog_epilog_substitution_excluded.pot').is_file()
+    pot = (app.outdir / 'prolog_epilog_substitution_excluded.pot').read_text(encoding='utf8')
+    msg_ids = list(filter(None, map(msgid_getter, pot.splitlines())))
+
+    assert msg_ids == [
+        "i18n without prologue and epilogue substitutions",
+        "This is content that does not include prologue and epilogue substitutions.",
+    ]