Add report functionality to the validation tool

validation/validate-mmcif-file.py

@@ -20,6 +20,7 @@ import argparse
 import atexit
 import copy
 import os
+import re
 import subprocess
 import sys
 import tempfile
@@ -85,6 +86,14 @@ def _parse_command_line():
         + "issues.",
         default=None,
     )
+    parser.add_argument(
+        "--report",
+        "-r",
+        action="store_true",
+        help="Print a concise report. Skips redundancies, may loose some "
+        + "information. Usable to verify your own writer on single files. "
+        + "Resolving problems of the report very often fixes the whole file.",
+    )
     parser.add_argument(
         "--verbose",
         "-v",
@@ -691,6 +700,127 @@ class _CifCheck:
                 msgs["cifcheck-errors"]
             )
 
+    def make_report(self):
+        """Make a concise report out of the results.
+
+        Be aware, that cuts away the majority of the messages. But solving those
+        issues first, may already repair a mmCIF file."""
+        not_implemented = ["errors"]
+        for category in not_implemented:
+            if len(self.check_results[category]) > 0:
+                raise NotImplementedError(
+                    f"Results for category '{category}' not yet supported in "
+                    + "report."
+                )
+        print("Report")
+        print("======")
+        print(f"Status of check: {self.check_results['status']}")
+        print("CIF dictionaries used:")
+        for dct in self.check_results["versions"]:
+            print(f"   {dct['title']}/ {dct['version']}")
+            print(f"   {dct['location']}")
+
+        # condense diagnosis data
+        rprt = {
+            "missing_cats": set(),
+            "missing_itms": set(),
+            "parchild_mm": set(),
+        }
+        for line in self.check_results["diagnosis"]:
+            # missing categories
+            for pttrn in [
+                r"^ERROR - category \"(?P<cat>.*)\" is mandatory, but it is "
+                + r"not present in datablock \"(?P<dblock>.*)\"$",
+                r"^\+\+ERROR - In block \"(?P<dblock>.*)\", parent category "
+                + r"\"(?P<cat>.*)\", of category \".*\", is missing\.$",
+            ]:
+                match = re.match(pttrn, line)
+                if match is not None:
+                    rprt["missing_cats"].add(match.group("cat"))
+                    _check_dblock_name(match.group("dblock"), rprt)
+                    break
+            if match is not None:
+                continue
+            # missing items
+            for pttrn in [
+                r"^ERROR - In block \"(?P<dblock>.*)\", mandatory "
+                + r"item \"(?P<itm>.*)\" is not in category \"(?P<cat>.*)\"$"
+            ]:
+                match = re.match(pttrn, line)
+                if match is not None:
+                    rprt["missing_itms"].add(
+                        f"{match.group('cat')}.{match.group('itm')}"
+                    )
+                    _check_dblock_name(match.group("dblock"), rprt)
+                    break
+            if match is not None:
+                continue
+            # parent-child issues
+            match = re.match(
+                r"^ERROR PARCHILD \".*\" - In block \"(?P<dblock>.*)\", in "
+                + r"category \"(?P<chld>.*)\", in row\# \d+, unmatched value "
+                + r"in the parent \"(?P<prnt>.*)\"$",
+                line,
+            )
+            if match is not None:
+                rprt["parchild_mm"].add(
+                    f"{match.group('chld')}->{match.group('prnt')}"
+                )
+                _check_dblock_name(match.group("dblock"), rprt)
+                continue
+            match = re.match(
+                r"^\"(?P<chld>.*)\" -> \"(?P<prnt>.*)\" value =(?P<vle>.*)$",
+                line,
+            )
+            if match is not None:
+                # prepare a string to be removed from parchild_mm
+                chld = match.group("chld").split(".")[0][1:]
+                prnt = match.group("prnt").split(".")[0][1:]
+                try:
+                    rprt["parchild_mm"].remove(f"{chld}->{prnt}")
+                except KeyError:
+                    pass
+                # add a more verbose line instead
+                rprt["parchild_mm"].add(
+                    f"{match.group('chld')}->{match.group('prnt')}, "
+                    + f"value={match.group('vle')}"
+                )
+                continue
+            # Unmatched lines need to be added to above evaluation
+            raise RuntimeError(
+                f'Unmatched diagnosis line found:\n"""{line}"""'
+            )
+
+        # print above evaluation in the report
+        # datablock
+        print("Diagnosis:")
+        print("   Datablock/ entry name:", rprt["datablock"])
+        print("   Missing categories:")
+        if len(rprt["missing_cats"]) > 0:
+            for line in sorted(rprt["missing_cats"]):
+                print(f"      {line}")
+        if len(rprt["missing_itms"]) > 0:
+            print("   Missing items:")
+            for line in sorted(rprt["missing_itms"]):
+                print(f"      {line}")
+        if len(rprt["parchild_mm"]) > 0:
+            print("   Mismatching parent/ child relationships:")
+            for line in sorted(rprt["parchild_mm"]):
+                print(f"      {line}")
+
+
+def _check_dblock_name(name, report):
+    """Compare datablock names."""
+    try:
+        # pylint: disable=used-before-assignment
+        if report["datablock"] != name:
+            raise RuntimeError(
+                "Two different datablock (names) found: "
+                + f"{report['datablock']} vs {name}"
+            )
+    except KeyError:
+        report["datablock"] = name
+
 
 def _find_utf(line):
     """Try to find a word with an UTF character in a string."""
@@ -773,6 +903,10 @@ def _main():
         print(f"Writing results of CifCheck to '{opts.out_file}'")
     cifcheck.make_json_output()
 
+    # print a report to stdout
+    if opts.report:
+        cifcheck.make_report()
+
     if cifcheck.got_issues():
         # If CifCheck found issues with the mmCIF file, exit with code 2. Exit
         # code 1 is reserved for general issues running the command, like "file