Spaces:
Running
Running
burtenshaw
commited on
Commit
·
a947fe6
1
Parent(s):
76563a1
switch to bs pdf command not decktape
Browse files- app/app.py +34 -86
app/app.py
CHANGED
|
@@ -136,115 +136,63 @@ def reconstruct_presentation_markdown(slides_data):
|
|
| 136 |
|
| 137 |
|
| 138 |
def generate_pdf_from_markdown(markdown_file_path, output_pdf_path):
|
| 139 |
-
"""Generates a PDF from a Markdown file using bs
|
| 140 |
-
logger.info(
|
|
|
|
|
|
|
| 141 |
working_dir = os.path.dirname(markdown_file_path)
|
| 142 |
markdown_filename = os.path.basename(markdown_file_path)
|
| 143 |
-
|
| 144 |
-
|
| 145 |
-
expected_html_filename = os.path.splitext(markdown_filename)[0] + ".html"
|
| 146 |
-
generated_html_path_abs = os.path.join(html_output_dir_abs, expected_html_filename)
|
| 147 |
-
pdf_gen_success = False # Flag to track success
|
| 148 |
-
|
| 149 |
-
# ---- Step 1: Generate HTML using bs export ----
|
| 150 |
-
try:
|
| 151 |
-
Path(html_output_dir_abs).mkdir(parents=True, exist_ok=True)
|
| 152 |
-
export_command = ["bs", "export", markdown_filename, "-o", html_output_dir_name]
|
| 153 |
-
logger.info(f"Running: {' '.join(export_command)} in CWD: {working_dir}")
|
| 154 |
-
export_result = subprocess.run(
|
| 155 |
-
export_command,
|
| 156 |
-
cwd=working_dir,
|
| 157 |
-
capture_output=True,
|
| 158 |
-
text=True,
|
| 159 |
-
check=True,
|
| 160 |
-
timeout=60,
|
| 161 |
-
)
|
| 162 |
-
logger.info("Backslide (bs export) OK.")
|
| 163 |
-
logger.debug(f"bs export stdout:\n{export_result.stdout}")
|
| 164 |
-
logger.debug(f"bs export stderr:\n{export_result.stderr}")
|
| 165 |
|
| 166 |
-
|
| 167 |
-
logger.error(f"Expected HTML not found: {generated_html_path_abs}")
|
| 168 |
-
try:
|
| 169 |
-
files_in_dir = os.listdir(html_output_dir_abs)
|
| 170 |
-
logger.error(f"Files in {html_output_dir_abs}: {files_in_dir}")
|
| 171 |
-
except FileNotFoundError:
|
| 172 |
-
logger.error(
|
| 173 |
-
f"HTML output directory {html_output_dir_abs} not found after bs run."
|
| 174 |
-
)
|
| 175 |
-
raise FileNotFoundError(
|
| 176 |
-
f"Generated HTML not found: {generated_html_path_abs}"
|
| 177 |
-
)
|
| 178 |
|
| 179 |
-
|
| 180 |
-
logger.error(
|
| 181 |
-
"`bs` command not found. Install backslide (`npm install -g backslide`)."
|
| 182 |
-
)
|
| 183 |
-
raise gr.Error("HTML generation tool (backslide/bs) not found.")
|
| 184 |
-
except subprocess.CalledProcessError as e:
|
| 185 |
-
logger.error(f"Backslide (bs export) failed (code {e.returncode}).")
|
| 186 |
-
logger.error(f"bs stderr:\n{e.stderr}")
|
| 187 |
-
raise gr.Error(f"Backslide HTML failed: {e.stderr[:500]}...")
|
| 188 |
-
except subprocess.TimeoutExpired:
|
| 189 |
-
logger.error("Backslide (bs export) timed out.")
|
| 190 |
-
raise gr.Error("HTML generation timed out (backslide).")
|
| 191 |
-
except Exception as e:
|
| 192 |
-
logger.error(f"Unexpected error during bs export: {e}", exc_info=True)
|
| 193 |
-
raise gr.Error(f"Unexpected error during HTML generation: {e}")
|
| 194 |
-
|
| 195 |
-
# ---- Step 2: Generate PDF from HTML using decktape ----
|
| 196 |
try:
|
| 197 |
-
|
| 198 |
-
|
| 199 |
-
|
| 200 |
-
|
| 201 |
-
|
| 202 |
-
|
| 203 |
capture_output=True,
|
| 204 |
text=True,
|
| 205 |
check=True,
|
| 206 |
-
timeout=
|
| 207 |
)
|
| 208 |
-
logger.info("
|
| 209 |
-
logger.debug(f"
|
| 210 |
-
logger.debug(f"
|
| 211 |
|
| 212 |
if os.path.exists(output_pdf_path):
|
| 213 |
logger.info(f"PDF generated successfully: {output_pdf_path}")
|
| 214 |
-
pdf_gen_success = True
|
| 215 |
return output_pdf_path
|
| 216 |
else:
|
| 217 |
-
logger.error("
|
| 218 |
-
|
| 219 |
|
| 220 |
except FileNotFoundError:
|
| 221 |
logger.error(
|
| 222 |
-
"`
|
|
|
|
|
|
|
|
|
|
|
|
|
| 223 |
)
|
| 224 |
-
raise gr.Error("PDF generation tool (decktape) not found.")
|
| 225 |
except subprocess.CalledProcessError as e:
|
| 226 |
-
logger.error(f"
|
| 227 |
-
|
| 228 |
-
|
|
|
|
| 229 |
except subprocess.TimeoutExpired:
|
| 230 |
-
logger.error("
|
| 231 |
-
raise gr.Error("PDF generation timed out (
|
| 232 |
except Exception as e:
|
| 233 |
-
logger.error(
|
| 234 |
-
f"Unexpected error during decktape PDF generation: {e}", exc_info=True
|
| 235 |
-
)
|
| 236 |
raise gr.Error(f"Unexpected error during PDF generation: {e}")
|
| 237 |
finally:
|
| 238 |
-
#
|
| 239 |
-
if os.path.exists(html_output_dir_abs):
|
| 240 |
-
try:
|
| 241 |
-
shutil.rmtree(html_output_dir_abs)
|
| 242 |
-
logger.info(f"Cleaned up HTML temp dir: {html_output_dir_abs}")
|
| 243 |
-
except Exception as cleanup_e:
|
| 244 |
-
logger.warning(
|
| 245 |
-
f"Could not cleanup HTML dir {html_output_dir_abs}: {cleanup_e}"
|
| 246 |
-
)
|
| 247 |
-
# Log final status
|
| 248 |
if pdf_gen_success:
|
| 249 |
logger.info(f"PDF generation process completed for {output_pdf_path}.")
|
| 250 |
else:
|
|
|
|
| 136 |
|
| 137 |
|
| 138 |
def generate_pdf_from_markdown(markdown_file_path, output_pdf_path):
|
| 139 |
+
"""Generates a PDF from a Markdown file using bs pdf."""
|
| 140 |
+
logger.info(
|
| 141 |
+
f"Attempting PDF gen using bs pdf: {markdown_file_path} -> {output_pdf_path}"
|
| 142 |
+
)
|
| 143 |
working_dir = os.path.dirname(markdown_file_path)
|
| 144 |
markdown_filename = os.path.basename(markdown_file_path)
|
| 145 |
+
# Ensure output directory exists (bs pdf might not create it)
|
| 146 |
+
Path(output_pdf_path).parent.mkdir(parents=True, exist_ok=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 147 |
|
| 148 |
+
pdf_gen_success = False
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 149 |
|
| 150 |
+
# ---- Run bs pdf ----
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 151 |
try:
|
| 152 |
+
# Use the absolute path for the output file
|
| 153 |
+
pdf_command = ["bs", "pdf", markdown_filename, "-o", str(output_pdf_path)]
|
| 154 |
+
logger.info(f"Running: {' '.join(pdf_command)} in CWD: {working_dir}")
|
| 155 |
+
pdf_result = subprocess.run(
|
| 156 |
+
pdf_command,
|
| 157 |
+
cwd=working_dir,
|
| 158 |
capture_output=True,
|
| 159 |
text=True,
|
| 160 |
check=True,
|
| 161 |
+
timeout=180, # Combined timeout (adjust as needed)
|
| 162 |
)
|
| 163 |
+
logger.info("bs pdf command executed successfully.")
|
| 164 |
+
logger.debug(f"bs pdf stdout:\n{pdf_result.stdout}")
|
| 165 |
+
logger.debug(f"bs pdf stderr:\n{pdf_result.stderr}")
|
| 166 |
|
| 167 |
if os.path.exists(output_pdf_path):
|
| 168 |
logger.info(f"PDF generated successfully: {output_pdf_path}")
|
| 169 |
+
pdf_gen_success = True
|
| 170 |
return output_pdf_path
|
| 171 |
else:
|
| 172 |
+
logger.error("bs pdf command finished but output PDF not found.")
|
| 173 |
+
raise gr.Error("bs pdf finished but failed to create the PDF file.")
|
| 174 |
|
| 175 |
except FileNotFoundError:
|
| 176 |
logger.error(
|
| 177 |
+
"`bs` command not found. Install backslide (`npm install -g backslide`) "
|
| 178 |
+
"and ensure its dependencies (like decktape) are met."
|
| 179 |
+
)
|
| 180 |
+
raise gr.Error(
|
| 181 |
+
"PDF generation tool (backslide/bs) not found or not configured correctly."
|
| 182 |
)
|
|
|
|
| 183 |
except subprocess.CalledProcessError as e:
|
| 184 |
+
logger.error(f"Backslide (bs pdf) failed (code {e.returncode}).")
|
| 185 |
+
# stderr from bs pdf might include decktape errors if it failed
|
| 186 |
+
logger.error(f"bs pdf stderr:\n{e.stderr}")
|
| 187 |
+
raise gr.Error(f"Backslide PDF generation failed: {e.stderr[:500]}...")
|
| 188 |
except subprocess.TimeoutExpired:
|
| 189 |
+
logger.error("Backslide (bs pdf) timed out.")
|
| 190 |
+
raise gr.Error("PDF generation timed out (backslide/bs pdf).")
|
| 191 |
except Exception as e:
|
| 192 |
+
logger.error(f"Unexpected error during bs pdf: {e}", exc_info=True)
|
|
|
|
|
|
|
| 193 |
raise gr.Error(f"Unexpected error during PDF generation: {e}")
|
| 194 |
finally:
|
| 195 |
+
# No intermediate HTML cleanup needed anymore
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 196 |
if pdf_gen_success:
|
| 197 |
logger.info(f"PDF generation process completed for {output_pdf_path}.")
|
| 198 |
else:
|