qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-10-06 22:50:14 +02:00
Code Issues Projects Releases Wiki Activity

Merge pull request #180 from bertsky/prepare-release-v0.5.0-fixlogging

Browse source 
prepare release v0.5.0: fix logging
This commit is contained in:
Konstantin Baierer 2025-09-24 12:28:10 +02:00 • committed by GitHub
commit 9c129c7f54
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 17 additions and 34 deletions
Download patch file Download diff file Expand all files Collapse all files

49
src/eynollah/eynollah.py
View file

@ -4531,6 +4531,21 @@ class Eynollah:
self.logger.debug("enter run") self.logger.debug("enter run")
t0_tot = time.time() t0_tot = time.time()
# Log enabled features directly
enabled_modes = []
if self.light_version:
enabled_modes.append("Light version")
if self.textline_light:
enabled_modes.append("Light textline detection")
if self.full_layout:
enabled_modes.append("Full layout analysis")
if self.ocr:
enabled_modes.append("OCR")
if self.tables:
enabled_modes.append("Table detection")
if enabled_modes:
self.logger.info("Enabled modes: " + ", ".join(enabled_modes))
if dir_in: if dir_in:
self.ls_imgs = os.listdir(dir_in) self.ls_imgs = os.listdir(dir_in)
self.ls_imgs = [ind_img for ind_img in self.ls_imgs if ind_img.endswith('.jpg') or ind_img.endswith('.jpeg') or ind_img.endswith('.png') or ind_img.endswith('.tif') or ind_img.endswith('.tiff') or ind_img.endswith('.JPG') or ind_img.endswith('.JPEG') or ind_img.endswith('.TIF') or ind_img.endswith('.TIFF') or ind_img.endswith('.PNG')] self.ls_imgs = [ind_img for ind_img in self.ls_imgs if ind_img.endswith('.jpg') or ind_img.endswith('.jpeg') or ind_img.endswith('.png') or ind_img.endswith('.tif') or ind_img.endswith('.tiff') or ind_img.endswith('.JPG') or ind_img.endswith('.JPEG') or ind_img.endswith('.TIF') or ind_img.endswith('.TIFF') or ind_img.endswith('.PNG')]
@ -4540,7 +4555,6 @@ class Eynollah:
raise ValueError("run requires either a single image filename or a directory") raise ValueError("run requires either a single image filename or a directory")
for img_filename in self.ls_imgs: for img_filename in self.ls_imgs:
print(img_filename, 'img_filename')
self.logger.info(img_filename) self.logger.info(img_filename)
t0 = time.time() t0 = time.time()
@ -4564,24 +4578,6 @@ class Eynollah:
t0 = time.time() t0 = time.time()
self.logger.info(f"Processing file: {self.writer.image_filename}") self.logger.info(f"Processing file: {self.writer.image_filename}")
# Log enabled features directly
enabled_modes = []
if self.light_version:
enabled_modes.append("Light version")
if self.textline_light:
enabled_modes.append("Light textline detection")
if self.full_layout:
enabled_modes.append("Full layout analysis")
if self.ocr:
enabled_modes.append("OCR")
if self.tables:
enabled_modes.append("Table detection")
if enabled_modes:
self.logger.info("Enabled modes: " + ", ".join(enabled_modes))
self.logger.info("Step 1/5: Image Enhancement") self.logger.info("Step 1/5: Image Enhancement")
img_res, is_image_enhanced, num_col_classifier, num_column_is_classified = self.run_enhancement(self.light_version) img_res, is_image_enhanced, num_col_classifier, num_column_is_classified = self.run_enhancement(self.light_version)
@ -5091,7 +5087,7 @@ class Eynollah:
t_order = time.time() t_order = time.time()
if self.full_layout: if self.full_layout:
self.logger.info(ep 4/5: Reading Order Detection") self.logger.info("Step 4/5: Reading Order Detection")
if self.reading_order_machine_based: if self.reading_order_machine_based:
self.logger.info("Using machine-based detection") self.logger.info("Using machine-based detection")
@ -5175,19 +5171,6 @@ class Eynollah:
all_found_textline_polygons_marginals_left, all_found_textline_polygons_marginals_right, all_box_coord_marginals_left, all_box_coord_marginals_right, slopes, slopes_h, slopes_marginals_left, slopes_marginals_right, all_found_textline_polygons_marginals_left, all_found_textline_polygons_marginals_right, all_box_coord_marginals_left, all_box_coord_marginals_right, slopes, slopes_h, slopes_marginals_left, slopes_marginals_right,
cont_page, polygons_lines_xml, ocr_all_textlines, ocr_all_textlines_h, ocr_all_textlines_marginals_left, ocr_all_textlines_marginals_right, ocr_all_textlines_drop, conf_contours_textregions, conf_contours_textregions_h) cont_page, polygons_lines_xml, ocr_all_textlines, ocr_all_textlines_h, ocr_all_textlines_marginals_left, ocr_all_textlines_marginals_right, ocr_all_textlines_drop, conf_contours_textregions, conf_contours_textregions_h)
summary = [
f"Total processing time: {time.time() - t0:.1f}s",
f"Output file: {self.writer.output_filename}"
]
if self.ocr:
summary.append("OCR processing completed")
if self.full_layout:
summary.append("Full layout analysis completed")
if self.tables:
summary.append("Table detection completed")
self.logger.info(f"Summary: {summary}")
return pcgts return pcgts
contours_only_text_parent_h = None contours_only_text_parent_h = None