dockerfile fix, txt path in api test fix

Author: sinkudo

tests/Dockerfile

@@ -4,8 +4,9 @@ FROM dedocproject/dedoc_p3.9_base:version_2023_08_28
 ENV PYTHONPATH "${PYTHONPATH}:/dedoc_root:/dedoc_root/tests:/dedoc_root/langchain"
 
 COPY requirements.txt .
-RUN pip3 install --no-cache-dir -r requirements.txt
 RUN pip3 install "langchain-community<1.0"
+RUN pip3 install --no-cache-dir -r requirements.txt
+RUN apt-get update && apt-get install -y --fix-missing --no-install-recommends fontforge
 
 RUN mkdir /dedoc_root
 COPY docs/source/_static/code_examples/langchain /dedoc_root/langchain

tests/api_tests/test_api_format_pdf_broken_encoding_reader.py

@@ -12,7 +12,6 @@ def _get_abs_path(self, file_name: str) -> str:
 
     def test_text_extraction(self) -> None:
         file_name = "mongolo.pdf"
-        orig_path = "../data/txt/mongolo.txt"
         result = self._send_request(file_name, dict(pdf_with_text_layer="bad_encoding_reader"))
         tree = result["content"]["structure"]
         text1 = self._get_by_tree_path(tree, "0.0")["text"]
@@ -22,6 +21,6 @@ def test_text_extraction(self) -> None:
         text5 = "\n".join(self._get_by_tree_path(tree, "0.1.2.0")["text"].split("\n")[:3])
 
         fulltext = text1 + text2 + text3 + text4 + text5
-        with open(orig_path, encoding="utf8", mode="r") as txt:
+        with open(os.path.join(self.data_directory_path, "txt", "mongolo.txt"), encoding="utf8", mode="r") as txt:
             accuracy = Levenshtein.ratio(txt.read(), fulltext)
             self.assertTrue(accuracy > 0.7)