Support documents with Archives

Allow "archives" parameters in Document creation.

Author: JorjMcKie

docs/pymupdf4llm/api.rst

@@ -279,6 +279,29 @@ The PyMuPDF4LLM API
             }
         }    
 
+.. method:: markdown_to_pdf(md_path: str | pathlib.Path, \
+    user_css: str | None = None, \
+    page_rect: rect-like | None = None, \
+    margins: rect-like | None) = None, \
+    archive: str | pathlib.Path | None = None, \
+    output_path: str | pathlib.Path | None = None) -> pymupdf.Document | None
+
+    Convert the markdown text content of the file specified by `md_path` into a PDF document.
+
+    The function is always available -- independently of whether you are using the PyMuPDF Layout module or not.
+
+    :arg str|Path md_path: the file path of the markdown file to be converted.
+
+    :arg str|None user_css: optional, a string of CSS code to be applied to the markdown content. This may be used to customize the appearance of the generated PDF document. If `None` (default), the built-in default CSS is used.
+
+    :arg rect-like|None page_rect: optional, the rectangle defining the page boundaries for the generated PDF document. If `None` (default), ISO A4 page dimensions are used. To use one of PyMuPDF's predefined page formats, use e.g. ``pymupdf.paper_rect("Letter")``.
+
+    :arg rect-like|None margins: optional, the margins (borders) for the generated pages. This must be a sequence of four floats ``[left, top, right, bottom]`` specifying the respective border width in points (1/72 inches). If `None` (default), the default ``[50, 50, 50, 50]`` margins are used.
+
+    :arg str|Archive|None archive: optional. This is be required if the markdown source references images that are **not** stored in the same folder as the markdown file. In this case, `archive` must be a `pymupdf.Archive` object which provides access to the respective image files. If `None` (default), it is assumed that all referenced images are stored in the same folder as the markdown file. The parameter **may** also be required if a custom ``user_css`` references external resources like font files.
+
+    :arg str|Path|None output_path: optional, the file path where the generated PDF document will be saved. If specified, the generated PDF will be saved to that location. If `None` (default), the document is returned as a `pymupdf.Document` object.
+
 .. note::
 
     Please see `this site <https://github.com/pymupdf/pymupdf4llm/discussions/327>`_ for more background and the current status of further improvements regarding usage with :ref:`PyMuPDF Layout <pymupdf-layout>`.

src/__init__.py

@@ -2897,7 +2897,7 @@ def __getitem__(self, i=0):
             raise IndexError(f"page {i} not in document")
         return self.load_page(i)
 
-    def __init__(self, filename=None, stream=None, filetype=None, rect=None, width=0, height=0, fontsize=11):
+    def __init__(self, filename=None, stream=None, filetype=None, archive=None, rect=None, width=0, height=0, fontsize=11):
         """Creates a document. Use 'open' as a synonym.
 
         Notes:
@@ -2943,7 +2943,16 @@ def __init__(self, filename=None, stream=None, filetype=None, rect=None, width=0
 
             self._name = filename
             self.stream = stream
-            
+            if isinstance(archive, pathlib.Path):
+                archive = Archive(archive.name)
+            elif isinstance(archive, str):
+                archive = Archive(archive)
+            if archive and not isinstance(archive, Archive):
+                raise TypeError(f"bad archive: {type(archive)=}.")
+            if archive:
+                archive_parm = archive.this  # pass this to open
+            else:
+                archive_parm = None  # means: no archive present
             if stream is not None:
                 if filename is not None and filetype is None:
                     # 2025-05-06: Use <filename> as the filetype. This is
@@ -2958,6 +2967,8 @@ def __init__(self, filename=None, stream=None, filetype=None, rect=None, width=0
                     stream = stream.getvalue()
                 else:
                     raise TypeError(f"bad stream: {type(stream)=}.")
+
+                # this prevents bad things if original goes out of existence:
                 self.stream = stream
 
                 assert isinstance(stream, (bytes, memoryview))
@@ -2967,9 +2978,9 @@ def __init__(self, filename=None, stream=None, filetype=None, rect=None, width=0
                     # raise a specific exception.
                     raise EmptyFileError('Cannot open empty stream.')
 
-                stream2 = mupdf.fz_open_memory(mupdf.python_buffer_data(stream), len(stream))
+                fz_stream = mupdf.fz_open_memory(mupdf.python_buffer_data(stream), len(stream))
                 try:
-                    doc = mupdf.fz_open_document_with_stream(filetype if filetype else '', stream2)
+                    doc = mupdf.fz_open_document_with_stream_and_dir(filetype if filetype else '', fz_stream, archive_parm)
                 except Exception as e:
                     if g_exceptions_verbose > 1:    exception_info()
                     raise FileDataError('Failed to open stream') from e
@@ -2996,20 +3007,15 @@ def __init__(self, filename=None, stream=None, filetype=None, rect=None, width=0
                     raise EmptyFileError(f'Cannot open empty file: {filename=}.')
 
                 if filetype:
-                    # Override the type implied by <filename>. MuPDF does not
-                    # have a way to do this directly so we open via a stream.
-                    try:
-                        fz_stream = mupdf.fz_open_file(filename)
-                        doc = mupdf.fz_open_document_with_stream(filetype, fz_stream)
-                    except Exception as e:
-                        if g_exceptions_verbose > 1:    exception_info()
-                        raise FileDataError(f'Failed to open file {filename!r} as type {filetype!r}.') from e
+                    suffix = filetype
                 else:
-                    try:
-                        doc = mupdf.fz_open_document(filename)
-                    except Exception as e:
-                        if g_exceptions_verbose > 1:    exception_info()
-                        raise FileDataError(f'Failed to open file {filename!r}.') from e
+                    suffix = Path(filename).suffix
+                try:
+                    fz_stream = mupdf.fz_open_file(filename)
+                    doc = mupdf.fz_open_document_with_stream_and_dir(suffix, fz_stream, archive_parm)
+                except Exception as e:
+                    if g_exceptions_verbose > 1:    exception_info()
+                    raise FileDataError(f'Failed to open file {filename!r} as type {suffix}.') from e
 
             else:
                 pdf = mupdf.PdfDocument()