diff --git a/pdfmajor/document/PDFDocument.py b/pdfmajor/document/PDFDocument.py index f853f72..41399b0 100644 --- a/pdfmajor/document/PDFDocument.py +++ b/pdfmajor/document/PDFDocument.py @@ -1,7 +1,8 @@ from pathlib import Path -from pdfmajor.document.PDFParsingContext import PDFParsingContext -from typing import BinaryIO +from typing import BinaryIO, List +from pdfmajor.document.pages import PDFPageTreeNode +from pdfmajor.document.PDFParsingContext import PDFParsingContext from pdfmajor.streambuffer import BufferStream @@ -24,6 +25,7 @@ class PDFDocument: self.__parser = PDFParsingContext(buffer) self.catalog = self.__parser.get_catalog() self.info = self.__parser.get_info() + self.pages: List[PDFPageTreeNode] = [] @property def health_report(self): diff --git a/pdfmajor/document/PDFDocumentCatalog.py b/pdfmajor/document/PDFDocumentCatalog.py index 9da5f8e..493b341 100644 --- a/pdfmajor/document/PDFDocumentCatalog.py +++ b/pdfmajor/document/PDFDocumentCatalog.py @@ -1,7 +1,7 @@ -from typing import Any, Dict, List, NamedTuple, Optional +from typing import Any, Dict, NamedTuple, Optional +from pdfmajor.document.pages import PDFPageTreeNode from pdfmajor.parser.objects.collections import PDFDictionary -from pdfmajor.parser.objects.indirect import ObjectRef from pdfmajor.parser.objects.primitives import PDFName @@ -9,7 +9,7 @@ class PDFDocumentCatalog(NamedTuple): """A Parsed version of the PDF Catalog object (see PDF spec 1.7 section 7.7.2 for more detail)""" version: Optional[PDFName] - pages: List[ObjectRef] + pages: PDFPageTreeNode page_labels: Optional[PDFDictionary] page_layout: Optional[PDFName] page_mode: Optional[PDFName] diff --git a/pdfmajor/document/PDFParsingContext.py b/pdfmajor/document/PDFParsingContext.py index dc446cb..54ca8cb 100644 --- a/pdfmajor/document/PDFParsingContext.py +++ b/pdfmajor/document/PDFParsingContext.py @@ -1,9 +1,10 @@ from pathlib import Path from typing import Any, BinaryIO, Dict, Optional, Set, Tuple, Type, TypeVar +from pdfmajor.document.pages import PDFPageTreeNode from pdfmajor.healthlog import PDFHealthReport from pdfmajor.parser.objects.base import PDFObject -from pdfmajor.parser.objects.collections import PDFArray, PDFDictionary +from pdfmajor.parser.objects.collections import PDFDictionary from pdfmajor.parser.objects.indirect import IndirectObject, ObjectRef from pdfmajor.parser.objects.primitives import PDFName from pdfmajor.streambuffer import BufferStream @@ -34,6 +35,14 @@ class PDFParsingContext: self.xrefdb = XRefDB(buffer) def get_object_from_ref(self, objref: ObjectRef) -> IndirectObject: + """Get an indirect object from it's object reference + + Args: + objref (ObjectRef) + + Returns: + IndirectObject + """ with self.buffer.get_window(): return self.xrefdb.get_obj( objref.obj_num, objref.gen_num, buffer=self.buffer @@ -126,20 +135,18 @@ class PDFParsingContext: pages_obj = cat_obj["Pages"] except KeyError: raise InvalidCatalogObj(f"Missing Pages entry {cat_obj}") - pages = [] if isinstance(pages_obj, ObjectRef): - pages = [pages_obj] - elif not isinstance(pages_obj, PDFArray): - raise InvalidCatalogObj(f"Invalid pages entry {pages_obj}") + pages = self.get_object_from_ref(pages_obj).get_object() + if not isinstance(pages, PDFDictionary): + raise InvalidCatalogObj( + f"Invalid pages ref entry {pages} from {pages_obj}" + ) + elif isinstance(pages_obj, PDFDictionary): + pages = pages_obj else: - for page in pages_obj: - if isinstance(page, ObjectRef): - pages.append(page) - else: - self.health_report.write( - "InvalidPageRef", f"The page {page} has an invalid value" - ) - validated_fields["pages"] = pages + raise InvalidCatalogObj(f"Invalid pages entry {pages_obj}") + + validated_fields["pages"] = PDFPageTreeNode.from_pdfdict(pages) validated_fields["raw"] = cat_obj return PDFDocumentCatalog(**validated_fields) diff --git a/pdfmajor/document/exceptions.py b/pdfmajor/document/exceptions.py index dbe657c..7a82dd7 100644 --- a/pdfmajor/document/exceptions.py +++ b/pdfmajor/document/exceptions.py @@ -31,3 +31,7 @@ class TooManyRectField(BrokenFilePDF): def __init__(self, values: List[Decimal]) -> None: super().__init__(f"Too many elements found for rectangle | values={values}") self.values = values + + +class InvalidPagesNodeKids(BrokenFilePDF): + pass diff --git a/pdfmajor/document/pages.py b/pdfmajor/document/pages.py index 08da89b..40e8b7d 100644 --- a/pdfmajor/document/pages.py +++ b/pdfmajor/document/pages.py @@ -1,13 +1,45 @@ +from decimal import Decimal from typing import List, NamedTuple, Optional + +from pdfmajor.document.exceptions import InvalidPagesNodeKids +from pdfmajor.document.structures import PDFRectangle +from pdfmajor.parser.objects.collections import PDFArray, PDFDictionary from pdfmajor.parser.objects.indirect import ObjectRef -from pdfmajor.parser.objects.collections import PDFDictionary +from pdfmajor.parser.objects.primitives import PDFInteger +from pdfmajor.parser.stream.PDFStream import PDFStream +from pdfmajor.util import validate_object_or_none class PDFPageTreeNode(NamedTuple): """A PDF Pages Node representation as it confirms with PDF spec 1.7 section 7.7.3.2""" kids: List[ObjectRef] - parent: Optional[ObjectRef] = None + parent: Optional[ObjectRef] + leaft_count: int + + raw: PDFDictionary + + @classmethod + def from_pdfdict(cls, pdfdict: PDFDictionary): + parent = validate_object_or_none(pdfdict.get("Parent"), ObjectRef) + kids = pdfdict["Kids"] + count = validate_object_or_none(pdfdict.get("Count"), PDFInteger) + vetted_kids: List[ObjectRef] = [] + if isinstance(kids, ObjectRef): + vetted_kids = [kids] + elif isinstance(kids, PDFArray): + for kid in kids: + vetted_kid = validate_object_or_none(kid, ObjectRef) + if vetted_kid: + vetted_kids.append(vetted_kid) + else: + raise InvalidPagesNodeKids(f"{kids}") + return cls( + kids=vetted_kids, + parent=parent, + raw=pdfdict, + leaft_count=0 if count is None else count.to_python(), + ) class PDFPage(NamedTuple): @@ -15,3 +47,11 @@ class PDFPage(NamedTuple): parent: ObjectRef resources: PDFDictionary + mediabox: PDFRectangle + bleedbox: PDFRectangle + trimbox: PDFRectangle + contents: List[PDFStream] + boxcolor_info: Optional[PDFDictionary] + rotate: Decimal + metadata: Optional[ObjectRef] + user_unit: Decimal diff --git a/pdfmajor/document/structures.py b/pdfmajor/document/structures.py index 37e338b..a12ae6e 100644 --- a/pdfmajor/document/structures.py +++ b/pdfmajor/document/structures.py @@ -1,8 +1,9 @@ from decimal import Decimal -from pdfmajor.document.exceptions import BrokenFilePDF, TooManyRectField -from pdfmajor.parser.objects.primitives import PDFInteger, PDFReal from typing import List, NamedTuple + +from pdfmajor.document.exceptions import BrokenFilePDF, TooManyRectField from pdfmajor.parser.objects.collections import PDFArray +from pdfmajor.parser.objects.primitives import PDFInteger, PDFReal class PDFRectangle(NamedTuple): diff --git a/pdfmajor/parser/objects/collections.py b/pdfmajor/parser/objects/collections.py index be250e2..efe203a 100644 --- a/pdfmajor/parser/objects/collections.py +++ b/pdfmajor/parser/objects/collections.py @@ -31,6 +31,9 @@ class PDFDictionary(Dict[str, PDFObject], PDFContextualObject): self.last_name: Optional[PDFName] = None self.strict = strict + def __repr__(self) -> str: + return f"PDFDictionary({self.to_python()})" + @classmethod def from_dict(cls, d: Dict[str, PDFObject]): new_obj = cls() diff --git a/pdfmajor/xref/xrefdb.py b/pdfmajor/xref/xrefdb.py index 03126b9..efc5a74 100644 --- a/pdfmajor/xref/xrefdb.py +++ b/pdfmajor/xref/xrefdb.py @@ -107,6 +107,19 @@ class XRefDB: gen_num: GenNum, buffer: BufferStream, ) -> IndirectObject: + """Get's an indirect object from either the pdf stream or a cached version in memeory + + Args: + obj_num (ObjNum) + gen_num (GenNum) + buffer (BufferStream) + + Raises: + InvalidXref: if the reference returned was not an indirect object + + Returns: + IndirectObject + """ raw_ref, xref = (obj_num, gen_num), self.xrefs[(obj_num, gen_num)] obj = self.objs.get(raw_ref, None) if obj is None: diff --git a/tests/test_document.py b/tests/test_document.py index da04d4c..26a8431 100644 --- a/tests/test_document.py +++ b/tests/test_document.py @@ -1,13 +1,14 @@ from decimal import Decimal -from pdfmajor.document.exceptions import BrokenFilePDF, TooManyRectField from typing import List -from pdfmajor.parser.objects.base import PDFObject -from pdfmajor.parser.objects.collections import PDFArray -from pdfmajor.document.structures import PDFRectangle from unittest import TestCase +from pdfmajor.document.exceptions import BrokenFilePDF, TooManyRectField +from pdfmajor.document.pages import PDFPageTreeNode from pdfmajor.document.PDFDocumentCatalog import PDFDocumentCatalog from pdfmajor.document.PDFParsingContext import PDFParsingContext +from pdfmajor.document.structures import PDFRectangle +from pdfmajor.parser.objects.base import PDFObject +from pdfmajor.parser.objects.collections import PDFArray, PDFDictionary from pdfmajor.parser.objects.indirect import ObjectRef from pdfmajor.parser.objects.primitives import ( PDFBoolean, @@ -23,8 +24,10 @@ class ParsingState(TestCase): def run_test(self, file_name: str, expcat: PDFDocumentCatalog): with all_pdf_files[file_name].get_window() as buffer: pdf = PDFParsingContext(buffer) + self.assertEqual(len(pdf.health_report), 0) + cat = pdf.get_catalog() self.assertEqual( - pdf.get_catalog(), + cat, expcat, ) @@ -33,21 +36,41 @@ class ParsingState(TestCase): "bad-unicode.pdf", PDFDocumentCatalog( version=None, - pages=[ObjectRef(141, 0)], + pages=PDFPageTreeNode.from_pdfdict( + PDFDictionary.from_dict( + { + "Count": PDFInteger(47, 0, 0), + "Type": PDFName("Pages", 0, 0), + "Kids": PDFArray.from_list( + [ + ObjectRef(142, 0), + ObjectRef(143, 0), + ObjectRef(144, 0), + ObjectRef(145, 0), + ObjectRef(146, 0), + ] + ), + } + ) + ), page_labels={"Nums": [PDFInteger(0, 0, 0), ObjectRef(140, 0)]}, page_layout=None, page_mode=None, - metadata={ - "Subtype": PDFName("XML", 0, 0), - "Length": PDFInteger(3529, 0, 0), - "Type": PDFName("Metadata", 0, 0), - }, - raw={ - "Metadata": ObjectRef(147, 0), - "Pages": ObjectRef(141, 0), - "Type": PDFName("Catalog", 0, 0), - "PageLabels": ObjectRef(139, 0), - }, + metadata=PDFDictionary.from_dict( + { + "Subtype": PDFName("XML", 0, 0), + "Length": PDFInteger(3529, 0, 0), + "Type": PDFName("Metadata", 0, 0), + } + ), + raw=PDFDictionary.from_dict( + { + "Metadata": ObjectRef(147, 0), + "Pages": ObjectRef(141, 0), + "Type": PDFName("Catalog", 0, 0), + "PageLabels": ObjectRef(139, 0), + } + ), ), ) @@ -56,24 +79,47 @@ class ParsingState(TestCase): "bar-charts.pdf", PDFDocumentCatalog( version=None, - pages=[ObjectRef(2, 0)], + pages=PDFPageTreeNode.from_pdfdict( + PDFDictionary.from_dict( + { + "Count": PDFInteger(6, 0, 0), + "Type": PDFName("Pages", 0, 0), + "Kids": PDFArray( + [ + ObjectRef(3, 0), + ObjectRef(9, 0), + ObjectRef(13, 0), + ObjectRef(36, 0), + ObjectRef(66, 0), + ObjectRef(188, 0), + ] + ), + } + ) + ), page_labels=None, page_layout=None, page_mode=None, - metadata={ - "Type": PDFName("Metadata", 0, 0), - "Subtype": PDFName("XML", 0, 0), - "Length": PDFInteger(3065, 0, 0), - }, - raw={ - "Type": PDFName("Catalog", 0, 0), - "Pages": ObjectRef(2, 0), - "Lang": PDFString(b"en-US", 0, 0), - "StructTreeRoot": ObjectRef(283, 0), - "MarkInfo": {"Marked": PDFBoolean(True, 0, 0)}, - "Metadata": ObjectRef(646, 0), - "ViewerPreferences": ObjectRef(647, 0), - }, + metadata=PDFDictionary.from_dict( + { + "Type": PDFName("Metadata", 0, 0), + "Subtype": PDFName("XML", 0, 0), + "Length": PDFInteger(3065, 0, 0), + } + ), + raw=PDFDictionary.from_dict( + { + "Type": PDFName("Catalog", 0, 0), + "Pages": ObjectRef(2, 0), + "Lang": PDFString(b"en-US", 0, 0), + "StructTreeRoot": ObjectRef(283, 0), + "MarkInfo": PDFDictionary.from_dict( + {"Marked": PDFBoolean(True, 0, 0)} + ), + "Metadata": ObjectRef(646, 0), + "ViewerPreferences": ObjectRef(647, 0), + } + ), ), )