add process_tei.py script, adjust some scripts where needed

README.md

@@ -93,9 +93,9 @@ In order to apply the workflow to any other publication copy it into the `runtim
 		$ ./scripts/run.py                 # run if not yet running
 		$ ./scripts/exec_in_container.py   # enter container
 
-1. eoatex -> pdf
+1. process eoatex:
 
-		$ process_eoa_latex.py -f input/example/124_eoatex/EOASample.tex
+		$ process_eoa_latex.py input/example/124_eoatex
 
 	(adjust filename if necessary)
 
@@ -111,26 +111,9 @@ In order to apply the workflow to any other publication copy it into the `input/
 		$ ./scripts/run.py                 # run if not yet running
 		$ ./scripts/exec_in_container.py   # enter container
 
-1. eoaTEI -> eoaTEI with bibliography
+1. process tei
 
-		$ tei_add_bibl.py input/example/125_tei_part
-
-1. eoaTEI -> eoaTEX
-
-		$ tei2eoatex.py input/example/125_tei_part
-
-1. eoaTEX -> pdf
-
-		$ eoatex2pdf.py -o output/125_tei_part/pdf output/125_tei_part/eoatex
-
-1. eoaTEI -> imxml (to intermediate xml)
-
-		$ gather_pickledata.py -o output/125_tei_part/pickle output/with_bibl/125_tei_part/{tei_part_with_bibl.xml,texfiles/example.bib}
-		tei2imxml.py -f tei_part.xml output/with_bibl/125_tei_part
-
-1. eoaTEI -> html
-
-		$ tei2html.py output/with_bibl/125_tei_part
+		$ process_tei.py input/example/125_tei_part
 
 ## The DocX workflow (DocX -> TEI -> ...) (TODO: describe how)
 

src/process_eoa_latex.py

@@ -49,13 +49,6 @@ def main(
     parser = argparse.ArgumentParser(
             formatter_class=argparse.ArgumentDefaultsHelpFormatter
     )
-    '''
-    parser.add_argument(
-            "-f", "--filename",
-            required = True,
-            help="Name of main EOATeX file (without suffix!)."
-    )
-    '''
     parser.add_argument(
             "-c", "--config",
             default = BASE_DIR / "config" / "eoaconvert.cfg",

src/process_tei.py

@@ -0,0 +1,92 @@
+#!/usr/bin/env python3
+
+from utils.load_config import load_config, exec_command
+
+# imports
+import argparse
+from pathlib import Path
+import glob
+import os
+import subprocess
+import shutil
+import logging
+
+BASE_DIR = Path( __file__ ).resolve().parent
+SCRIPT_PATH = Path( __file__ )
+SCRIPT_NAME = SCRIPT_PATH.stem
+
+DEFAULT_INPUT_DIR = \
+    Path(os.environ['INPUT_DIR'] if 'INPUT_DIR' in os.environ else './input')
+
+DEFAULT_OUTPUT_DIR = \
+    Path(os.environ['OUTPUT_DIR'] if 'OUTPUT_DIR' in os.environ else './output')
+
+DEFAULT_DEPENDENCIES_DIR = \
+    Path(os.environ['DEPENDENCIES_DIR'] if 'DEPENDENCIES_DIR' in os.environ else './dependencies')
+
+def main(
+    publ_dir
+):
+    PUBL_NAME = publ_dir.resolve().stem
+    exec_command(
+        f"tei_add_bibl.py -! \"{publ_dir}\""
+    )
+    exec_command(
+        f"tei2eoatex.py -! \"{publ_dir}\""
+    )
+    exec_command(
+        f"eoatex2pdf.py --output-dir \"{DEFAULT_OUTPUT_DIR}/{PUBL_NAME}/pdf\" \"{DEFAULT_OUTPUT_DIR}/{PUBL_NAME}/eoatex\""
+    )
+    exec_command(
+        f"tei_pickle.py -! \"{DEFAULT_OUTPUT_DIR}/with_bibl/{PUBL_NAME}\""
+    )
+    exec_command(
+        f"tei2imxml.py --filename no_bibl.xml \"{DEFAULT_OUTPUT_DIR}/with_bibl/{PUBL_NAME}\""
+    )
+    exec_command(
+        f"tei2html.py -! --filename \"with_bibl.xml\" \"{DEFAULT_OUTPUT_DIR}/with_bibl/{PUBL_NAME}\""
+    )
+
+if __name__ == "__main__":
+
+    #####################
+    # Parsing arguments #
+    #####################
+
+    parser = argparse.ArgumentParser(
+            formatter_class=argparse.ArgumentDefaultsHelpFormatter
+    )
+    parser.add_argument(
+            "-c", "--config",
+            default = BASE_DIR / "config" / "eoaconvert.cfg",
+            dest="CONFIG_FILE",
+            help="Name of configuration file",
+            metavar="CONFIGURATION"
+    )
+    parser.add_argument(
+            "-l", "--log-file",
+            default = (DEFAULT_OUTPUT_DIR / 'logs' / SCRIPT_NAME).with_suffix(".log"),
+            help="logfile"
+    )
+    parser.add_argument(
+            "--log-level",
+            default = "INFO",
+            help="log level: choose between DEBUG, INFO, WARNING, ERROR, CRITICAL"
+    )
+    parser.add_argument(
+        "PUBLICATION_DIR",
+        help = "directory containing the publication (including resources like pictures, etc.)",
+        type = Path,
+)
+
+    args = parser.parse_args()
+
+    CONFIG = load_config(
+            args.CONFIG_FILE,
+            args.log_level,
+            args.log_file,
+    )
+
+    main(
+            publ_dir = Path( args.PUBLICATION_DIR )
+    )

src/tei2html.py

@@ -70,12 +70,14 @@ def copy_dir(
             default = "INFO",
             help="log level: choose between DEBUG, INFO, WARNING, ERROR, CRITICAL"
     )
+    '''
     parser.add_argument(
             "--root-dir",
             default = DEFAULT_OUTPUT_DIR / "html_from_tei",
             type = Path,
             help=""
     )
+    '''
     parser.add_argument(
             "-p", "--param",
             action = 'append',
@@ -91,7 +93,7 @@ def copy_dir(
     )
     parser.add_argument(
             "-f", "--filename",
-            default = Path("*_with_bibl.xml"),
+            default = Path("with_bibl.xml"),
             type = Path,
             help = "xml file inside PUBLICATION_DIR, or absolute path. Patterns like '*.xml' are also acceptable"
     )
@@ -150,11 +152,20 @@ def copy_dir(
     output_dir = \
         args.output_dir if args.output_dir is not None else (DEFAULT_OUTPUT_DIR / publ_dir.resolve().stem) / "html"
 
-    CONFIG_FILE = args.config
+    if not tei_filename.is_file():
+        raise( Exception(
+            f"not a valid input file: {tei_filename}"
+        ) )
+
+    if output_dir.exists():
+        if args.overwrite:
+            rmtree( output_dir )
+        else:
+            raise( Exception( f"output directory already existing: '{output_dir}'!" ) )
 
+    CONFIG_FILE = args.config
     log_dir = output_dir / "log"
     log_file = (log_dir / SCRIPT_NAME) . with_suffix( ".log" )
-
     CONFIG = load_config(
             CONFIG_FILE,
             args.log_level,
@@ -163,16 +174,6 @@ def copy_dir(
 
     logging.info( f"tei_file: {tei_filename}, publ_dir: {publ_dir}" )
 
-    if not tei_filename.is_file():
-        raise( Exception(
-            f"not a valid input file: {tei_filename}"
-        ) )
-
-    if output_dir.exists():
-        if args.overwrite:
-            rmtree( output_dir )
-        else:
-            raise( Exception( f"output directory already existing: '{output_dir}'!" ) )
     if not output_dir.exists():
         mkdir( output_dir )
         ## copy webdesign:

src/tei_add_bibl.py

@@ -220,3 +220,24 @@ def create_bibl(
         )
     else:
         raise( Exception("unknown publication type!"))
+
+    # create uniquely named links to
+    # original tei file and the one with added bibliography
+    orig_link = output_dir / "no_bibl.xml"
+    if orig_link.exists():
+        logging.error(
+                f"error while creating unique link: file already exists '{orig_link}'"
+        )
+        exit(1)
+    orig_link . symlink_to(
+            tei_file . name
+    )
+    with_bibl_link = output_dir / "with_bibl.xml"
+    if with_bibl_link.exists():
+        logging.error(
+                f"error while creating unique link: file already exists '{with_bibl_link}'"
+        )
+        exit(1)
+    with_bibl_link . symlink_to(
+            tei_with_bibl_file . name
+    )