Merge pull request #36 from SAFEHR-data/paul/fix-test-imports

Fix unit tests for the analysis module

Author: p-j-smith

analysis/README.md src/pyonb/analysis/README.mdanalysis/README.md renamed to src/pyonb/analysis/README.md

@@ -1,6 +1,6 @@
 # OCR performance evaluation metrics
 
-The `analysis` directory contains code for evaluating the performance of OCR tools. The metrics include:
+The `pyonb.analysis` module contains code for evaluating the performance of OCR tools. The metrics include:
 
 - Character Error Rate (CER)
 - Word Error Rate (WER)
@@ -14,7 +14,7 @@ For all of these metrics 0 = perfect score and 1 = maximum dissimilarity.
 
 2. **OCR Processed File:** Process your PDF file with an OCR tool. Either save the output as a .txt file, or in the case of pyonb, save it as a .json file (e.g. `ocr-output.json`). Note, the JSON response from pyonb is a dictionary structure which contains a key-value pair `"ocr-result": "example output text"`
 
-3. Run OCR performance metrics (code below uses files in [tests/](../tests/data/ocr_eval/) directory):
+3. Run OCR performance metrics (code below uses files in [tests/](../../../tests/data/ocr_eval/) directory):
 
 ```shell
 cd pyonb/

analysis/__init__.py src/pyonb/analysis/__init__.pyanalysis/__init__.py renamed to src/pyonb/analysis/__init__.py

@@ -5,7 +5,7 @@
 from ast import literal_eval
 from pathlib import Path
 
-from analysis.metrics import cer, ned, wer
+from pyonb.analysis.metrics import cer, ned, wer
 
 
 def read_file(file_path: Path) -> str | dict:

analysis/eval_ocr.py src/pyonb/analysis/eval_ocr.pyanalysis/eval_ocr.py renamed to src/pyonb/analysis/eval_ocr.py

@@ -3,7 +3,7 @@
 from ast import literal_eval
 from pathlib import Path
 
-from analysis.eval_ocr import evaluate_metrics, read_file
+from pyonb.analysis.eval_ocr import evaluate_metrics, read_file
 
 
 def test_read_file(ground_truth_txt_filepath: Path, marker_ocr_json_filepath: Path) -> None: