from-pdf-pt now uses *.pdf as default glob; prepared for release

waikato-llm · Jul 4, 2024 · 5b63994 · 5b63994
1 parent c93bf71
commit 5b63994
Show file tree

Hide file tree

Showing 3 changed files with 9 additions and 3 deletions.
diff --git a/CHANGES.rst b/CHANGES.rst
@@ -1,6 +1,12 @@
 Changelog
 =========
 
+0.0.2 (2024-07-05)
+------------------
+
+- `from-pdf-pt` now uses `*.pdf` as default glob
+
+
 0.0.1 (2024-05-02)
 ------------------
 

diff --git a/setup.py b/setup.py
@@ -31,11 +31,11 @@ def _read(f):
     },
     packages=find_namespace_packages(where='src'),
     install_requires=[
-        "llm-dataset-converter>=0.2.1",
+        "llm-dataset-converter>=0.2.4",
         "pypdf",
         "simple-range>=0.0.3",
     ],
-    version="0.0.1",
+    version="0.0.2",
     author='Peter Reutemann',
     author_email='fracpete@waikato.ac.nz',
     entry_points={

diff --git a/src/ldc_pdf/pretrain/_pdf.py b/src/ldc_pdf/pretrain/_pdf.py
@@ -102,7 +102,7 @@ def initialize(self):
         Initializes the reading, e.g., for opening files or databases.
         """
         super().initialize()
-        self._inputs = locate_files(self.source, input_lists=self.source_list, fail_if_empty=True)
+        self._inputs = locate_files(self.source, input_lists=self.source_list, fail_if_empty=True, default_glob="*.pdf")
         if self.page_range is None:
             self.page_range = "first-last"
         if self.combine_pages is None: