From 4653b53a6c15d8450eccf6d68d6a6abd620ff401 Mon Sep 17 00:00:00 2001 From: Ari Sosnovsky Date: Tue, 27 Oct 2020 12:26:56 -0400 Subject: [PATCH] fixed stream parsing --- pdfmajor/parser/__init__.py | 17 ++++------ pdfmajor/parser/objects/indirect.py | 11 +++---- .../stream.py => stream/PDFStream.py} | 33 +++++++++---------- .../parser/{ => stream}/PDFStreamParser.py | 0 pdfmajor/parser/stream/__init__.py | 0 tests/test_parser.py | 32 +++++++++++------- 6 files changed, 47 insertions(+), 46 deletions(-) rename pdfmajor/parser/{objects/stream.py => stream/PDFStream.py} (79%) rename pdfmajor/parser/{ => stream}/PDFStreamParser.py (100%) create mode 100644 pdfmajor/parser/stream/__init__.py diff --git a/pdfmajor/parser/__init__.py b/pdfmajor/parser/__init__.py index 71b1a04..94a8906 100644 --- a/pdfmajor/parser/__init__.py +++ b/pdfmajor/parser/__init__.py @@ -1,3 +1,4 @@ +from pdfmajor.parser.stream.PDFStream import PDFStream from pdfmajor.parser.xref import XRefRow from pdfmajor.parser.xrefdb import XRefDB from typing import Iterator, Optional @@ -11,9 +12,8 @@ from pdfmajor.lexer.token import ( TokenKeyword, ) from pdfmajor.lexer import iter_tokens -from .objects.stream import PDFStream from .objects.indirect import IndirectObject -from .objects.base import PDFContextualObject, PDFObject +from .objects.base import PDFObject from .objects.collections import PDFArray, PDFDictionary from .objects.comment import PDFComment from .state import ParsingState @@ -101,7 +101,7 @@ def _on_endobj(state: ParsingState, token: TokenKeyword) -> IndirectObject: raise BrokenFile(f"Invalid number of integers in obj defition {token}") for num in state.flush_int_collections(): last_ctx.pass_item(num) - return last_ctx.into_readonly_copy() + return last_ctx else: raise ParserError(f"Recived an endobj while not procesing an indirect object") @@ -120,20 +120,17 @@ def _on_stream( last_ctx = state.current_context if last_ctx is not None and isinstance(last_ctx, IndirectObject): xrefdb.save_indobject(last_ctx) - stream = PDFStream(token.end_loc + 1, 0) + # stream = PDFStream(token.end_loc + 1, 0) obj = _save_or_get_object(buffer, xrefdb, last_ctx.obj_num, last_ctx.gen_num) - with buffer.get_window(): - next_token = next(iter_tokens(buffer)) + last_ctx.stream = stream = PDFStream.from_pdfdict(token.end_loc, obj) + buffer.seek(stream.offset + stream.length) + next_token = next(iter_tokens(buffer)) if not ( isinstance(next_token, TokenKeyword) and next_token.value == b"endstream" ): raise BrokenFile( f"stream did not contain 'endstream', instead the token {next_token} was found" ) - stream.pass_item(obj) - buffer.seek(stream.offset + stream.length) - last_ctx.save_object(stream) - state.context_stack.append(stream) else: raise ParserError(f"PDFStream is missing initilization dictionary") diff --git a/pdfmajor/parser/objects/indirect.py b/pdfmajor/parser/objects/indirect.py index dd1269b..3a9d9e5 100644 --- a/pdfmajor/parser/objects/indirect.py +++ b/pdfmajor/parser/objects/indirect.py @@ -1,7 +1,7 @@ from typing import Optional -from ..exceptions import InvalidIndirectObjAccess + +from pdfmajor.parser.stream.PDFStream import PDFStream from .base import PDFContextualObject, PDFObject -from .primitives import PDFNull class IndirectObject(PDFContextualObject): @@ -13,11 +13,13 @@ class IndirectObject(PDFContextualObject): gen_num: int, offset: int, data: Optional[PDFObject] = None, + stream: Optional[PDFStream] = None, ) -> None: self.offset = offset self.obj_num = obj_num self.gen_num = gen_num self.__data = data + self.stream = stream def __repr__(self) -> str: return f"IndirectObject(obj_num={self.obj_num}, gen_num={self.gen_num}, offset={self.offset}, data={self.get_object()})" @@ -34,11 +36,6 @@ class IndirectObject(PDFContextualObject): """ self.__data = obj - def into_readonly_copy(self): - return IndirectObject( - self.obj_num, self.gen_num, self.offset, self.get_object() - ) - def to_python(self): obj = self.get_object() if obj is None: diff --git a/pdfmajor/parser/objects/stream.py b/pdfmajor/parser/stream/PDFStream.py similarity index 79% rename from pdfmajor/parser/objects/stream.py rename to pdfmajor/parser/stream/PDFStream.py index 99a3900..54e237b 100644 --- a/pdfmajor/parser/objects/stream.py +++ b/pdfmajor/parser/stream/PDFStream.py @@ -1,11 +1,11 @@ from pdfmajor.parser.exceptions import ParserError from typing import Any, Dict, List, Optional, Union -from .base import PDFContextualObject, PDFObject -from .primitives import PDFName -from .collections import PDFDictionary + +from ..objects.base import PDFObject +from ..objects.collections import PDFDictionary -class PDFStream(PDFContextualObject): +class PDFStream: """A class representing a PDF Stream as specified in PDF spec 1.7 section 7.3.8""" def __init__( @@ -18,7 +18,6 @@ class PDFStream(PDFContextualObject): ffilter: Optional[List[str]] = None, fdecode_parms: Optional[List[Dict[str, Any]]] = None, dl: Optional[int] = None, - other_metadata: Optional[Dict[str, Any]] = None, ) -> None: """ Args: @@ -39,25 +38,23 @@ class PDFStream(PDFContextualObject): self.ffilter: Optional[List[str]] = ffilter self.fdecode_parms: Optional[List[Dict[str, Any]]] = fdecode_parms self.dl: Optional[int] = dl - self.other_metadata: Dict[str, Any] = ( - other_metadata if other_metadata is not None else {} - ) - def pass_item(self, item: PDFObject) -> None: + @classmethod + def from_pdfdict(cls, offset: int, item: PDFObject) -> None: if not isinstance(item, PDFDictionary): raise ParserError(f"Cannot initilize a pdfstream with {type(item)}") data = item.to_python() - self.length = data["Length"] - self.filter = get_single_or_list(data.get("Filter", self.filter)) - self.decode_parms = get_single_or_list( - data.get("DecodeParms", self.decode_parms) + stream = cls(offset, data["Length"]) + stream.filter = get_single_or_list(data.get("Filter", stream.filter)) + stream.decode_parms = get_single_or_list( + data.get("DecodeParms", stream.decode_parms) ) - self.f = data.get("F", self.f) - self.ffilter = get_single_or_list(data.get("FFilter", self.ffilter)) - self.fdecode_parms = get_single_or_list( - data.get("FDecodeParms", self.fdecode_parms) + stream.f = data.get("F", stream.f) + stream.ffilter = get_single_or_list(data.get("FFilter", stream.ffilter)) + stream.fdecode_parms = get_single_or_list( + data.get("FDecodeParms", stream.fdecode_parms) ) - self.other_metadata = data + return stream def to_python(self): return { diff --git a/pdfmajor/parser/PDFStreamParser.py b/pdfmajor/parser/stream/PDFStreamParser.py similarity index 100% rename from pdfmajor/parser/PDFStreamParser.py rename to pdfmajor/parser/stream/PDFStreamParser.py diff --git a/pdfmajor/parser/stream/__init__.py b/pdfmajor/parser/stream/__init__.py new file mode 100644 index 0000000..e69de29 diff --git a/tests/test_parser.py b/tests/test_parser.py index 9893037..e6248e7 100644 --- a/tests/test_parser.py +++ b/tests/test_parser.py @@ -12,7 +12,7 @@ from pdfmajor.streambuffer import BufferStream from pdfmajor.parser.objects.indirect import IndirectObject from pdfmajor.parser.objects.base import PDFObject from pdfmajor.parser.objects.comment import PDFComment -from pdfmajor.parser.objects.stream import PDFStream +from pdfmajor.parser.stream.PDFStream import PDFStream CURRENT_FOLDER = Path(__file__).parent @@ -184,7 +184,7 @@ class IndirectObjects(TestCase): buffer = BufferStream.from_bytes( br""" 5 0 obj -<< /Type /XRef /Length 124 /Filter /FlateDecode >> +<< /Type /XRef /Length 123 /Filter /FlateDecode /otherfeature [10 20]>> stream "<99>l01234567890l01234567890l01234567890l01234567890l01234567890l01234567890l01234567890l01234567890l01234567 <80>endstream @@ -193,18 +193,28 @@ endobj ) for obj in iter_objects(buffer, XRefDB()): self.assertIsInstance(obj, IndirectObject) - self.assertIsInstance(obj.get_object(), PDFStream) + self.assertIsInstance(obj.get_object(), PDFDictionary) + self.assertIsInstance(obj.stream, PDFStream) # type: ignore + self.assertDictEqual( + obj.stream.to_python(), # type: ignore + { + "offset": 99, + "length": 123, + "filter": ["/FlateDecode"], + "decode_parms": [], + "f": None, + "ffilter": [], + "fdecode_parms": [], + "dl": None, + }, + ) self.assertDictEqual( obj.get_object().to_python(), { - "offset": 78, - "length": 0, - "filter": None, - "decode_parms": None, - "f": None, - "ffilter": None, - "fdecode_parms": None, - "dl": None, + "Filter": "/FlateDecode", + "Length": 123, + "Type": "/XRef", + "otherfeature": [10, 20], }, )