diff --git a/pdfmajor/document/PDFDocument.py b/pdfmajor/document/PDFDocument.py index 5424243..7ffc6ec 100644 --- a/pdfmajor/document/PDFDocument.py +++ b/pdfmajor/document/PDFDocument.py @@ -1,7 +1,8 @@ from pathlib import Path -from typing import BinaryIO, List +from typing import BinaryIO, Iterator, List, Optional from pdfmajor.document.pages import PDFPage +from pdfmajor.document.parsers.pages import iter_all_page_leafs from pdfmajor.document.parsers.root import get_catalog, get_info from pdfmajor.document.PDFParsingContext import PDFParsingContext from pdfmajor.streambuffer import BufferStream @@ -24,11 +25,48 @@ class PDFDocument: def __init__(self, buffer: BufferStream) -> None: self.__parser = PDFParsingContext(buffer) + self.__processed_pages: Optional[List[PDFPage]] = None + self.__processed_pages_done = False self.catalog = get_catalog(self.__parser) self.info = get_info(self.__parser) - self.pages: List[PDFPage] = [] @property def health_report(self): """A log of errors that occured in the document during the processing of the file""" return self.__parser.health_report + + @property + def num_pages(self) -> int: + """Returns the number of pages that the metadata of the document claims it has. + This number should be trusted, unless you suspect the document may have encoutered some corruption. To get an excat count do `len(list(self.iter_pages()))`. Please note that this will load all of the metadata of each document into memory. + + Returns: + int + """ + return self.catalog.pages.leaft_count + + def iter_pages(self) -> Iterator[PDFPage]: + """Iterate over all of the pages in the document, this will also keep a record of the pdf page in memory for quicker future retrieval! (use self.clear() to free it up) + + Yields: + Iterator[PDFPage] + """ + if self.__processed_pages is not None and self.__processed_pages_done: + return iter(self.__processed_pages) + else: + self.__processed_pages = [] + for page in iter_all_page_leafs(self.__parser, self.catalog.pages): + self.__processed_pages.append(page) + yield page + self.__processed_pages_done = True + if len(self.__processed_pages) != self.num_pages: + print( + "WARN! number of pages proceseed with iter_pages did not match number of pages specified by the document's metadata!" + f"\n --> expected pages = {self.num_pages}" + f"\n --> found pages = {len(self.__processed_pages)}" + ) + + def clear(self): + """Clears out memory used by the class as much as possible""" + self.__processed_pages = None + self.__processed_pages_done = False diff --git a/pdfmajor/document/parsers/pages.py b/pdfmajor/document/parsers/pages.py index 98a845f..aea0bad 100644 --- a/pdfmajor/document/parsers/pages.py +++ b/pdfmajor/document/parsers/pages.py @@ -1,5 +1,5 @@ from decimal import Decimal -from typing import Any, Dict, List, Optional +from typing import Any, Dict, Iterator, List, Optional from pdfmajor.document.exceptions import BrokenFilePDF from pdfmajor.parser.objects.base import PDFObject @@ -16,10 +16,18 @@ from ..utils import iter_single_ref_as_array_ref PDFRawFields = Dict[str, Any] -def get_all_page_leafs( +def iter_all_page_leafs( pctx: PDFParsingContext, root_node: PDFPageTreeNode -) -> List[PDFPage]: - all_leafs: List[PDFPage] = [] +) -> Iterator[PDFPage]: + """Iterate over every page in the pdf + + Args: + pctx (PDFParsingContext) + root_node (PDFPageTreeNode) + + Yields: + Iterator[PDFPage] + """ current_root: List[PDFPageTreeNode] = [root_node] current_root_aspage: List[Dict[str, Optional[PDFObject]]] = [ parse_pdfpage_fields(pctx, root_node.raw, allow_missing=True) @@ -53,7 +61,7 @@ def get_all_page_leafs( current_child_idx.append(0) continue elif obj["Type"].to_python() == "/Page": - all_leafs.append(make_pdfpage(croot_pagedata, pctx, obj)) + yield make_pdfpage(croot_pagedata, pctx, obj) current_child_idx[-1] += 1 continue # catch all other cases (the above expression will issue a 'continue') @@ -62,7 +70,6 @@ def get_all_page_leafs( f"Expecting page ref {page_ref} to be a dictionary with type 'Pages' or 'Page', but got a {obj}", ) current_child_idx[-1] += 1 - return all_leafs def make_pdfpage( diff --git a/tests/test_document.py b/tests/test_document.py index f399163..5c05408 100644 --- a/tests/test_document.py +++ b/tests/test_document.py @@ -4,7 +4,7 @@ from unittest import TestCase from pdfmajor.document.exceptions import BrokenFilePDF, TooManyRectField from pdfmajor.document.pages import PDFPageTreeNode -from pdfmajor.document.parsers.pages import get_all_page_leafs +from pdfmajor.document.parsers.pages import iter_all_page_leafs from pdfmajor.document.parsers.root import get_catalog, get_info from pdfmajor.document.PDFDocument import PDFDocument from pdfmajor.document.PDFDocumentCatalog import PDFDocumentCatalog @@ -133,7 +133,7 @@ class ParsingState(TestCase): catalog = get_catalog(parser) self.assertEqual(len(parser.health_report), 0) get_info(parser) - get_all_page_leafs(parser, catalog.pages) + list(iter_all_page_leafs(parser, catalog.pages)) def test_doc(self): with all_pdf_files["bad-unicode.pdf"].get_window() as buffer: