cvfile 0.1.0__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (36) hide show
  1. cvfile-0.1.0/.gitignore +9 -0
  2. cvfile-0.1.0/PKG-INFO +74 -0
  3. cvfile-0.1.0/README.md +45 -0
  4. cvfile-0.1.0/examples/build_python_sample.py +70 -0
  5. cvfile-0.1.0/pyproject.toml +57 -0
  6. cvfile-0.1.0/src/cvfile/__init__.py +49 -0
  7. cvfile-0.1.0/src/cvfile/_constants.py +26 -0
  8. cvfile-0.1.0/src/cvfile/_pdf.py +255 -0
  9. cvfile-0.1.0/src/cvfile/_security.py +183 -0
  10. cvfile-0.1.0/src/cvfile/_types.py +89 -0
  11. cvfile-0.1.0/src/cvfile/_xmp.py +240 -0
  12. cvfile-0.1.0/src/cvfile/detect.py +25 -0
  13. cvfile-0.1.0/src/cvfile/embed/__init__.py +39 -0
  14. cvfile-0.1.0/src/cvfile/embed/_chunk.py +130 -0
  15. cvfile-0.1.0/src/cvfile/embed/_embed.py +66 -0
  16. cvfile-0.1.0/src/cvfile/embed/_embeddings.py +176 -0
  17. cvfile-0.1.0/src/cvfile/embed/_huggingface.py +88 -0
  18. cvfile-0.1.0/src/cvfile/embed/_search.py +78 -0
  19. cvfile-0.1.0/src/cvfile/extract.py +65 -0
  20. cvfile-0.1.0/src/cvfile/inspect.py +25 -0
  21. cvfile-0.1.0/src/cvfile/integrations/__init__.py +5 -0
  22. cvfile-0.1.0/src/cvfile/integrations/langchain.py +144 -0
  23. cvfile-0.1.0/src/cvfile/integrations/llamaindex.py +113 -0
  24. cvfile-0.1.0/src/cvfile/pack.py +173 -0
  25. cvfile-0.1.0/src/cvfile/server/__init__.py +31 -0
  26. cvfile-0.1.0/src/cvfile/server/_conneg.py +129 -0
  27. cvfile-0.1.0/src/cvfile/server/_handler.py +176 -0
  28. cvfile-0.1.0/src/cvfile/server/asgi.py +146 -0
  29. cvfile-0.1.0/src/cvfile/server/wsgi.py +125 -0
  30. cvfile-0.1.0/src/cvfile/validate.py +140 -0
  31. cvfile-0.1.0/tests/test_embed.py +134 -0
  32. cvfile-0.1.0/tests/test_interop.py +66 -0
  33. cvfile-0.1.0/tests/test_round_trip.py +105 -0
  34. cvfile-0.1.0/tests/test_security.py +60 -0
  35. cvfile-0.1.0/tests/test_server_conneg.py +74 -0
  36. cvfile-0.1.0/tests/test_server_handlers.py +195 -0
cvfile-0.1.0/.gitignore ADDED
@@ -0,0 +1,9 @@
1
+ .venv/
2
+ __pycache__/
3
+ *.pyc
4
+ .pytest_cache/
5
+ .mypy_cache/
6
+ .ruff_cache/
7
+ *.egg-info/
8
+ dist/
9
+ build/
cvfile-0.1.0/PKG-INFO ADDED
@@ -0,0 +1,74 @@
1
+ Metadata-Version: 2.4
2
+ Name: cvfile
3
+ Version: 0.1.0
4
+ Summary: Reference SDK for the .cv open file format — pack, extract, inspect, validate.
5
+ Project-URL: Homepage, https://cvfile.org
6
+ Project-URL: Repository, https://github.com/cvfile/cv
7
+ Author: cvfile.org
8
+ License: Apache-2.0
9
+ Keywords: ats,cv,embeddings,markdown,pdf,pdfa,rag,resume
10
+ Classifier: Development Status :: 3 - Alpha
11
+ Classifier: License :: OSI Approved :: Apache Software License
12
+ Classifier: Programming Language :: Python :: 3
13
+ Classifier: Programming Language :: Python :: 3.10
14
+ Classifier: Programming Language :: Python :: 3.11
15
+ Classifier: Programming Language :: Python :: 3.12
16
+ Classifier: Programming Language :: Python :: 3.13
17
+ Requires-Python: >=3.10
18
+ Requires-Dist: pypdf<6,>=5.0
19
+ Provides-Extra: dev
20
+ Requires-Dist: cbor2>=5.6; extra == 'dev'
21
+ Requires-Dist: httpx>=0.27; extra == 'dev'
22
+ Requires-Dist: mypy>=1.13; extra == 'dev'
23
+ Requires-Dist: pytest>=8.0; extra == 'dev'
24
+ Requires-Dist: ruff>=0.7; extra == 'dev'
25
+ Provides-Extra: embed
26
+ Requires-Dist: cbor2>=5.6; extra == 'embed'
27
+ Requires-Dist: httpx>=0.27; extra == 'embed'
28
+ Description-Content-Type: text/markdown
29
+
30
+ # `cvfile`
31
+
32
+ Reference Python SDK for the [`.cv`](https://cvfile.org) open file format.
33
+
34
+ ## Install
35
+
36
+ ```bash
37
+ pip install cvfile
38
+ ```
39
+
40
+ ## Pack
41
+
42
+ ```python
43
+ from cvfile import pack
44
+
45
+ with open("resume.pdf", "rb") as f:
46
+ pdf_bytes = f.read()
47
+ with open("resume.md") as f:
48
+ md = f.read()
49
+
50
+ cv_bytes = pack(
51
+ pdf=pdf_bytes,
52
+ markdown=md,
53
+ metadata={"primary_language": "en"},
54
+ )
55
+
56
+ with open("resume.cv", "wb") as f:
57
+ f.write(cv_bytes)
58
+ ```
59
+
60
+ ## Extract
61
+
62
+ ```python
63
+ from cvfile import extract, extract_markdown
64
+
65
+ file = extract(open("resume.cv", "rb").read())
66
+ print(file.metadata.version) # "0.1"
67
+ print([p.name for p in file.payloads]) # ['resume.md', 'resume.html']
68
+
69
+ md = extract_markdown(open("resume.cv", "rb").read())
70
+ ```
71
+
72
+ ## License
73
+
74
+ Apache-2.0.
cvfile-0.1.0/README.md ADDED
@@ -0,0 +1,45 @@
1
+ # `cvfile`
2
+
3
+ Reference Python SDK for the [`.cv`](https://cvfile.org) open file format.
4
+
5
+ ## Install
6
+
7
+ ```bash
8
+ pip install cvfile
9
+ ```
10
+
11
+ ## Pack
12
+
13
+ ```python
14
+ from cvfile import pack
15
+
16
+ with open("resume.pdf", "rb") as f:
17
+ pdf_bytes = f.read()
18
+ with open("resume.md") as f:
19
+ md = f.read()
20
+
21
+ cv_bytes = pack(
22
+ pdf=pdf_bytes,
23
+ markdown=md,
24
+ metadata={"primary_language": "en"},
25
+ )
26
+
27
+ with open("resume.cv", "wb") as f:
28
+ f.write(cv_bytes)
29
+ ```
30
+
31
+ ## Extract
32
+
33
+ ```python
34
+ from cvfile import extract, extract_markdown
35
+
36
+ file = extract(open("resume.cv", "rb").read())
37
+ print(file.metadata.version) # "0.1"
38
+ print([p.name for p in file.payloads]) # ['resume.md', 'resume.html']
39
+
40
+ md = extract_markdown(open("resume.cv", "rb").read())
41
+ ```
42
+
43
+ ## License
44
+
45
+ Apache-2.0.
cvfile-0.1.0/examples/build_python_sample.py ADDED
@@ -0,0 +1,70 @@
1
+ """Build a Python-produced .cv fixture for cross-SDK interop testing.
2
+
3
+ Outputs to packages/sdk-js/tests/fixtures/python-produced.cv so the JS test
4
+ suite can verify it extracts identically.
5
+ """
6
+
7
+ from __future__ import annotations
8
+
9
+ import io
10
+ from pathlib import Path
11
+
12
+ import pypdf
13
+
14
+ from cvfile import extract, inspect, pack, validate
15
+
16
+
17
+ SAMPLE_MD = """# Marie Curie
18
+
19
+ Physicist and chemist · Paris, France
20
+
21
+ ## Notable
22
+
23
+ * Discovered polonium and radium
24
+ * Two Nobel Prizes (Physics 1903, Chemistry 1911)
25
+ """
26
+
27
+ SAMPLE_HTML = """<!doctype html>
28
+ <html lang="en"><head><meta charset="utf-8"><title>Marie Curie</title></head>
29
+ <body><h1>Marie Curie</h1><p>Physicist and chemist.</p></body></html>"""
30
+
31
+
32
+ def make_blank_pdf() -> bytes:
33
+ writer = pypdf.PdfWriter()
34
+ writer.add_blank_page(width=300, height=400)
35
+ buf = io.BytesIO()
36
+ writer.write(buf)
37
+ return buf.getvalue()
38
+
39
+
40
+ def main() -> None:
41
+ repo_root = Path(__file__).resolve().parents[3]
42
+ out_dir = repo_root / "packages" / "sdk-js" / "tests" / "fixtures"
43
+ out_dir.mkdir(parents=True, exist_ok=True)
44
+
45
+ cv = pack(
46
+ pdf=make_blank_pdf(),
47
+ markdown=SAMPLE_MD,
48
+ html=SAMPLE_HTML,
49
+ json_resume={"basics": {"name": "Marie Curie"}},
50
+ metadata={"primary_language": "en", "generator": "cvfile-py-examples/marie-curie"},
51
+ )
52
+
53
+ out_path = out_dir / "python-produced.cv"
54
+ out_path.write_bytes(cv)
55
+ print(f"Wrote {out_path} ({len(cv)} bytes)")
56
+
57
+ # Self-verify the fixture is sane.
58
+ file = extract(cv)
59
+ print(f" payloads: {[p.name for p in file.payloads]}")
60
+ print(f" metadata: version={file.metadata.version} lang={file.metadata.primary_language}")
61
+
62
+ meta = inspect(cv)
63
+ assert meta.primary_payload == "resume.md"
64
+
65
+ report = validate(cv)
66
+ print(f" validate: ok={report.ok} issues={len(report.issues)}")
67
+
68
+
69
+ if __name__ == "__main__":
70
+ main()
cvfile-0.1.0/pyproject.toml ADDED
@@ -0,0 +1,57 @@
1
+ [build-system]
2
+ requires = ["hatchling"]
3
+ build-backend = "hatchling.build"
4
+
5
+ [project]
6
+ name = "cvfile"
7
+ version = "0.1.0"
8
+ description = "Reference SDK for the .cv open file format — pack, extract, inspect, validate."
9
+ readme = "README.md"
10
+ license = { text = "Apache-2.0" }
11
+ requires-python = ">=3.10"
12
+ authors = [{ name = "cvfile.org" }]
13
+ keywords = ["cv", "resume", "pdf", "pdfa", "markdown", "embeddings", "rag", "ats"]
14
+ classifiers = [
15
+ "Development Status :: 3 - Alpha",
16
+ "License :: OSI Approved :: Apache Software License",
17
+ "Programming Language :: Python :: 3",
18
+ "Programming Language :: Python :: 3.10",
19
+ "Programming Language :: Python :: 3.11",
20
+ "Programming Language :: Python :: 3.12",
21
+ "Programming Language :: Python :: 3.13",
22
+ ]
23
+ dependencies = [
24
+ "pypdf>=5.0,<6",
25
+ ]
26
+
27
+ [project.urls]
28
+ Homepage = "https://cvfile.org"
29
+ Repository = "https://github.com/cvfile/cv"
30
+
31
+ [project.optional-dependencies]
32
+ embed = [
33
+ "cbor2>=5.6",
34
+ "httpx>=0.27",
35
+ ]
36
+ dev = [
37
+ "pytest>=8.0",
38
+ "ruff>=0.7",
39
+ "mypy>=1.13",
40
+ "cbor2>=5.6",
41
+ "httpx>=0.27",
42
+ ]
43
+
44
+ [tool.hatch.build.targets.wheel]
45
+ packages = ["src/cvfile"]
46
+
47
+ [tool.ruff]
48
+ line-length = 120
49
+ target-version = "py310"
50
+
51
+ [tool.ruff.lint]
52
+ select = ["E", "F", "I", "B", "UP", "N", "SIM", "RUF"]
53
+
54
+ [tool.mypy]
55
+ python_version = "3.12"
56
+ strict = true
57
+ warn_unreachable = true
cvfile-0.1.0/src/cvfile/__init__.py ADDED
@@ -0,0 +1,49 @@
1
+ """Reference SDK for the .cv open file format."""
2
+
3
+ from cvfile._constants import (
4
+ CV_NAMESPACE_PREFIX,
5
+ CV_NAMESPACE_URI,
6
+ CV_SPEC_VERSION,
7
+ DEFAULT_PAYLOAD_NAMES,
8
+ PAYLOAD_MIME_TYPES,
9
+ )
10
+ from cvfile._types import (
11
+ AlternateMeta,
12
+ CvFile,
13
+ CvMetadata,
14
+ EmbeddingSpaceSummary,
15
+ ExtractedPayload,
16
+ IntegrityEntry,
17
+ Payload,
18
+ ValidationIssue,
19
+ ValidationReport,
20
+ )
21
+ from cvfile.detect import is_cv_file
22
+ from cvfile.extract import extract, extract_html, extract_markdown
23
+ from cvfile.inspect import inspect
24
+ from cvfile.pack import pack
25
+ from cvfile.validate import validate
26
+
27
+ __all__ = [
28
+ "CV_NAMESPACE_PREFIX",
29
+ "CV_NAMESPACE_URI",
30
+ "CV_SPEC_VERSION",
31
+ "DEFAULT_PAYLOAD_NAMES",
32
+ "PAYLOAD_MIME_TYPES",
33
+ "AlternateMeta",
34
+ "CvFile",
35
+ "CvMetadata",
36
+ "EmbeddingSpaceSummary",
37
+ "ExtractedPayload",
38
+ "IntegrityEntry",
39
+ "Payload",
40
+ "ValidationIssue",
41
+ "ValidationReport",
42
+ "extract",
43
+ "extract_html",
44
+ "extract_markdown",
45
+ "inspect",
46
+ "is_cv_file",
47
+ "pack",
48
+ "validate",
49
+ ]
cvfile-0.1.0/src/cvfile/_constants.py ADDED
@@ -0,0 +1,26 @@
1
+ """Constants shared across the cvfile SDK."""
2
+
3
+ CV_SPEC_VERSION = "0.1"
4
+
5
+ CV_NAMESPACE_URI = "http://ns.cvfile.org/cv/1.0/"
6
+ CV_NAMESPACE_PREFIX = "cv"
7
+
8
+ DEFAULT_GENERATOR = f"cvfile-py/{CV_SPEC_VERSION}"
9
+
10
+ DEFAULT_PAYLOAD_NAMES = {
11
+ "markdown": "resume.md",
12
+ "html": "resume.html",
13
+ "json": "resume.json",
14
+ "embeddings": "embeddings.cbor",
15
+ }
16
+
17
+ PAYLOAD_MIME_TYPES = {
18
+ "markdown": "text/markdown",
19
+ "html": "text/html",
20
+ "json": "application/json",
21
+ "embeddings": "application/vnd.cv.embeddings+cbor",
22
+ "pdf": "application/pdf",
23
+ "cv": "application/vnd.cv+pdf",
24
+ }
25
+
26
+ MAX_PAYLOAD_BYTES_DEFAULT = 16 * 1024 * 1024
cvfile-0.1.0/src/cvfile/_pdf.py ADDED
@@ -0,0 +1,255 @@
1
+ """Thin pypdf wrapper for /AF Associated Files and /Metadata streams.
2
+
3
+ This module isolates pypdf so we can swap to pikepdf later without changing the
4
+ public API. PDF/A-3 conformance work that requires deeper PDF rewriting (font
5
+ embedding, ICC profile injection on arbitrary input PDFs) will land here.
6
+ """
7
+
8
+ from __future__ import annotations
9
+
10
+ import io
11
+ from dataclasses import dataclass
12
+ from datetime import datetime, timezone
13
+ from typing import Literal
14
+
15
+ from pypdf import PdfReader, PdfWriter
16
+ from pypdf.generic import (
17
+ ArrayObject,
18
+ ByteStringObject,
19
+ DecodedStreamObject,
20
+ DictionaryObject,
21
+ IndirectObject,
22
+ NameObject,
23
+ NumberObject,
24
+ StreamObject,
25
+ TextStringObject,
26
+ )
27
+
28
+ AFRelationshipKind = Literal["Alternative", "Data", "Supplement"]
29
+
30
+
31
+ @dataclass(frozen=True, slots=True)
32
+ class RawPayload:
33
+ name: str
34
+ mime_type: str
35
+ relationship: AFRelationshipKind
36
+ bytes_: bytes
37
+ description: str | None = None
38
+
39
+
40
+ def load_writer(pdf_bytes: bytes) -> PdfWriter:
41
+ reader = PdfReader(io.BytesIO(pdf_bytes))
42
+ writer = PdfWriter(clone_from=reader)
43
+ return writer
44
+
45
+
46
+ def add_associated_file(
47
+ writer: PdfWriter,
48
+ *,
49
+ name: str,
50
+ data: bytes,
51
+ mime_type: str,
52
+ description: str,
53
+ relationship: AFRelationshipKind,
54
+ creation_date: datetime,
55
+ modification_date: datetime,
56
+ ) -> None:
57
+ """Attach `data` as an Associated File on the document catalog (/AF)."""
58
+
59
+ embedded_stream = DecodedStreamObject()
60
+ embedded_stream.set_data(data)
61
+ embedded_stream.update(
62
+ {
63
+ NameObject("/Type"): NameObject("/EmbeddedFile"),
64
+ NameObject("/Subtype"): NameObject(_mime_to_name(mime_type)),
65
+ NameObject("/Length"): NumberObject(len(data)),
66
+ NameObject("/Params"): DictionaryObject(
67
+ {
68
+ NameObject("/CreationDate"): TextStringObject(_pdf_date(creation_date)),
69
+ NameObject("/ModDate"): TextStringObject(_pdf_date(modification_date)),
70
+ NameObject("/Size"): NumberObject(len(data)),
71
+ }
72
+ ),
73
+ }
74
+ )
75
+ embedded_ref = writer._add_object(embedded_stream)
76
+
77
+ filespec = DictionaryObject(
78
+ {
79
+ NameObject("/Type"): NameObject("/Filespec"),
80
+ NameObject("/F"): TextStringObject(name),
81
+ NameObject("/UF"): TextStringObject(name),
82
+ NameObject("/Desc"): TextStringObject(description),
83
+ NameObject("/AFRelationship"): NameObject(f"/{relationship}"),
84
+ NameObject("/EF"): DictionaryObject(
85
+ {NameObject("/F"): embedded_ref, NameObject("/UF"): embedded_ref}
86
+ ),
87
+ }
88
+ )
89
+ filespec_ref = writer._add_object(filespec)
90
+
91
+ catalog = writer._root_object
92
+ af_array = catalog.get(NameObject("/AF"))
93
+ if isinstance(af_array, IndirectObject):
94
+ af_array = af_array.get_object()
95
+ if not isinstance(af_array, ArrayObject):
96
+ af_array = ArrayObject()
97
+ catalog[NameObject("/AF")] = af_array
98
+ af_array.append(filespec_ref)
99
+
100
+
101
+ def set_metadata_xml(writer: PdfWriter, xml: str) -> None:
102
+ data = xml.encode("utf-8")
103
+ stream = DecodedStreamObject()
104
+ stream.set_data(data)
105
+ stream.update(
106
+ {
107
+ NameObject("/Type"): NameObject("/Metadata"),
108
+ NameObject("/Subtype"): NameObject("/XML"),
109
+ NameObject("/Length"): NumberObject(len(data)),
110
+ }
111
+ )
112
+ ref = writer._add_object(stream)
113
+ writer._root_object[NameObject("/Metadata")] = ref
114
+
115
+
116
+ def write_to_bytes(writer: PdfWriter) -> bytes:
117
+ _ensure_trailer_id(writer)
118
+ buf = io.BytesIO()
119
+ writer.write(buf)
120
+ return buf.getvalue()
121
+
122
+
123
+ def _ensure_trailer_id(writer: PdfWriter) -> None:
124
+ """Set the trailer /ID array (PDF/A-3u rule 6.1.3)."""
125
+ import secrets
126
+
127
+ if getattr(writer, "_ID", None):
128
+ return
129
+ id_hex = secrets.token_hex(16).upper().encode("ascii")
130
+ writer._ID = ArrayObject([ByteStringObject(id_hex), ByteStringObject(id_hex)])
131
+
132
+
133
+ def read_associated_files(reader: PdfReader) -> list[RawPayload]:
134
+ catalog = reader.trailer.get("/Root")
135
+ if catalog is None:
136
+ return []
137
+ catalog = catalog.get_object() if isinstance(catalog, IndirectObject) else catalog
138
+ af = catalog.get("/AF") if isinstance(catalog, DictionaryObject) else None
139
+ if af is None:
140
+ return []
141
+ if isinstance(af, IndirectObject):
142
+ af = af.get_object()
143
+ if not isinstance(af, ArrayObject):
144
+ return []
145
+
146
+ out: list[RawPayload] = []
147
+ for entry in af:
148
+ filespec = entry.get_object() if isinstance(entry, IndirectObject) else entry
149
+ payload = _parse_filespec(filespec)
150
+ if payload:
151
+ out.append(payload)
152
+ return out
153
+
154
+
155
+ def read_metadata_xml(reader: PdfReader) -> str | None:
156
+ root = reader.trailer.get("/Root")
157
+ if root is None:
158
+ return None
159
+ root = root.get_object() if isinstance(root, IndirectObject) else root
160
+ meta = root.get("/Metadata") if isinstance(root, DictionaryObject) else None
161
+ if meta is None:
162
+ return None
163
+ meta = meta.get_object() if isinstance(meta, IndirectObject) else meta
164
+ if not isinstance(meta, StreamObject):
165
+ return None
166
+ data = meta.get_data()
167
+ if isinstance(data, str):
168
+ return data
169
+ return data.decode("utf-8", errors="replace")
170
+
171
+
172
+ def _parse_filespec(filespec: DictionaryObject) -> RawPayload | None:
173
+ if not isinstance(filespec, DictionaryObject):
174
+ return None
175
+ ef = filespec.get("/EF")
176
+ if ef is None:
177
+ return None
178
+ ef = ef.get_object() if isinstance(ef, IndirectObject) else ef
179
+ if not isinstance(ef, DictionaryObject):
180
+ return None
181
+ stream_ref = ef.get("/UF") or ef.get("/F")
182
+ if stream_ref is None:
183
+ return None
184
+ stream = stream_ref.get_object() if isinstance(stream_ref, IndirectObject) else stream_ref
185
+ if not isinstance(stream, StreamObject):
186
+ return None
187
+
188
+ data = stream.get_data()
189
+ if isinstance(data, str):
190
+ data = data.encode("latin-1", errors="replace")
191
+
192
+ name_obj = filespec.get("/UF") or filespec.get("/F")
193
+ if name_obj is None:
194
+ return None
195
+ name = str(name_obj)
196
+
197
+ subtype = stream.get("/Subtype") or filespec.get("/Subtype")
198
+ mime_type = _name_to_mime(str(subtype)) if subtype else "application/octet-stream"
199
+
200
+ desc_obj = filespec.get("/Desc")
201
+ description = str(desc_obj) if desc_obj else None
202
+
203
+ rel_obj = filespec.get("/AFRelationship")
204
+ rel_str = str(rel_obj).lstrip("/") if rel_obj else "Supplement"
205
+ if rel_str not in {"Alternative", "Data", "Supplement"}:
206
+ rel_str = "Supplement"
207
+ rel: AFRelationshipKind = rel_str # type: ignore[assignment]
208
+
209
+ return RawPayload(name=name, mime_type=mime_type, relationship=rel, bytes_=bytes(data), description=description)
210
+
211
+
212
+ def _mime_to_name(mime: str) -> str:
213
+ """Wrap a MIME type for use as a PDF Name. pypdf's NameObject handles
214
+ the per-character #XX escaping itself when serializing; we MUST NOT
215
+ pre-escape, otherwise the '#' of our own escape gets re-escaped to '#23'.
216
+ """
217
+ return "/" + mime
218
+
219
+
220
+ def _name_to_mime(name: str) -> str:
221
+ s = name.lstrip("/")
222
+ out = []
223
+ i = 0
224
+ while i < len(s):
225
+ c = s[i]
226
+ if c == "#" and i + 2 < len(s):
227
+ try:
228
+ out.append(chr(int(s[i + 1 : i + 3], 16)))
229
+ i += 3
230
+ continue
231
+ except ValueError:
232
+ pass
233
+ out.append(c)
234
+ i += 1
235
+ return "".join(out)
236
+
237
+
238
+ def _pdf_date(dt: datetime) -> str:
239
+ """Format a datetime as a PDF date string (D:YYYYMMDDHHmmSS+HH'mm')."""
240
+ if dt.tzinfo is None:
241
+ dt = dt.replace(tzinfo=timezone.utc)
242
+ return dt.astimezone(timezone.utc).strftime("D:%Y%m%d%H%M%SZ")
243
+
244
+
245
+ __all__ = [
246
+ "AFRelationshipKind",
247
+ "ByteStringObject",
248
+ "RawPayload",
249
+ "add_associated_file",
250
+ "load_writer",
251
+ "read_associated_files",
252
+ "read_metadata_xml",
253
+ "set_metadata_xml",
254
+ "write_to_bytes",
255
+ ]