You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@apr.apache.org by jo...@apache.org on 2026/08/06 14:38:48 UTC
svn commit: r1936931 - apr/site/trunk
Author: jorton
Date: Thu Aug 6 14:38:48 2026
New Revision: 1936931
Log:
Generate list of vulnerabilites from CVE*.json input at the
end of /security_report.html.
Assisted-by: Claude Sonnet 4.6 <no...@anthropic.com>
Added:
apr/site/trunk/cvejsontohtml.py (contents, props changed)
Modified:
apr/site/trunk/build.xml
Modified: apr/site/trunk/build.xml
==============================================================================
--- apr/site/trunk/build.xml Thu Aug 6 14:37:04 2026 (r1936930)
+++ apr/site/trunk/build.xml Thu Aug 6 14:38:48 2026 (r1936931)
@@ -31,6 +31,7 @@
<taskdef name="anakia" classname="org.apache.velocity.anakia.AnakiaTask">
<classpath refid="classpath"/>
</taskdef>
+ <delete file="${docs.dest}/security_report.html"/>
<anakia basedir="${docs.src}" destdir="${docs.dest}/"
extension=".html" style="./site.vsl"
projectFile="stylesheets/project.xml"
@@ -54,5 +55,23 @@
</fileset>
</copy>
+ <exec executable="python3"
+ output="${docs.dest}/security-cve-fragment.html"
+ failonerror="true">
+ <arg value="cvejsontohtml.py"/>
+ <arg value="-i"/>
+ <arg value="security-json/"/>
+ <arg value="-l"/>
+ <arg value="https://httpd.apache.org/security/impact_levels.html"/>
+ <arg value="-n"/>
+ </exec>
+ <loadfile property="cve.content"
+ srcFile="${docs.dest}/security-cve-fragment.html"/>
+ <property name="cve.content" value=""/>
+ <replace file="${docs.dest}/security_report.html"
+ token="</blockquote>"
+ value="</blockquote>${cve.content}"/>
+ <delete file="${docs.dest}/security-cve-fragment.html"/>
+
</target>
</project>
Added: apr/site/trunk/cvejsontohtml.py
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ apr/site/trunk/cvejsontohtml.py Thu Aug 6 14:38:48 2026 (r1936931)
@@ -0,0 +1,213 @@
+import json
+import os
+import re
+import xml.sax.saxutils as saxutils
+import sys
+import traceback
+from optparse import OptionParser
+parser = OptionParser()
+
+parser.add_option("-v","--version",help="major version to filter on",dest="filterversion")
+parser.add_option("-e","--extratext",help="extra text to add to description",dest="extratext")
+parser.add_option("-i","--inputdirectory",help="directory of json files",dest="directory")
+parser.add_option("-d","--debug",help="enable debug output to stderr",dest="debug",action="store_true",default=False)
+parser.add_option("-l","--impactlink",help="URL for impact levels page",dest="impactlink",default="/security/impact_levels.html")
+parser.add_option("-n","--nopreamble",help="skip the page header/preamble output",dest="nopreamble",action="store_true",default=False)
+
+# To debug without getting flooded, put 1 working and 1 broken JSON file in tmp/ and pass --debug -i tmp
+
+(options,args) = parser.parse_args()
+
+re_fixedin = re.compile('(released in )?(?P<released>\\d\\.\\d\\.\\d+(-\\S+)?)( released)?');
+
+def natural_sort_key(s, _nsre=re.compile('([0-9]+)')):
+ return [int(text) if text.isdigit() else text.lower()
+ for text in _nsre.split(s)]
+
+filterversion = options.filterversion or ""
+cves = []
+entries = {}
+DEFAULT_CVE_DATA_VERSION = "4.0" # Default (old) CVE data version
+
+for x in os.listdir(options.directory or "./"):
+ if x.endswith(".json"):
+ if options.debug:
+ print(f"Processing: {x}", file=sys.stderr)
+ try:
+ fd = open(options.directory+x)
+ cve = json.load(fd)
+ cve["_filename"] = x
+ cves.append(cve)
+ except:
+ print ("Ignoring due to error parsing: "+x)
+ if options.debug:
+ print(f"Skipped {x}: parse error", file=sys.stderr)
+ continue
+
+# Filter on version and store by release(s) that fixed it
+for cve in cves:
+ # Establish which version of CVE JSON we are dealing with
+ data_version = cve.get("dataVersion", DEFAULT_CVE_DATA_VERSION)
+ #print("%s: v%s" % (cve["_filename"], data_version), file=sys.stderr)
+
+ if data_version == DEFAULT_CVE_DATA_VERSION: # Old style CVE
+ cve["id"] = cve["CVE_data_meta"]["ID"]
+ timearray = cve["timeline"]
+ elif data_version.startswith("5"): # Newer style JSON
+ cve["id"] = cve["cveMetadata"]["cveId"]
+ timearray = cve["containers"]["cna"]["timeline"]
+ else:
+ print(f"unknown data version {data_version} in cve {cve['_filename']}", file=sys.stderr)
+ sys.exit(1)
+
+ if options.debug:
+ print(f"Filtering: {cve['id']} from {cve['_filename']}", file=sys.stderr)
+
+ matched = False
+ for time in timearray:
+ timed = time["value"]
+ matcher = re_fixedin.match(timed);
+ if (matcher and matcher.group('released').startswith(filterversion)):
+ fixedin = matcher.group('released')
+ if (not fixedin in entries):
+ entries[fixedin] = []
+ entries[fixedin].append(cve)
+ matched = True
+
+ if options.debug and not matched:
+ print(f"Skipped {cve['id']}: no matching version for filter '{filterversion}'", file=sys.stderr)
+
+# We want to sort on reverse number fixed, where our versions are dotted numbers, except some special cases
+# like never-fixed, or -dev fixed which should always appear first
+# Then we want to sort in the section by CVE name, handling numerics correctly
+
+lastfixedv = ""
+productname = ""
+sections = []
+for k,v in sorted(entries.items(), key=lambda s: [int(u) if u.isdigit() else 999 for u in s[0].split(',')[0].split('.')], reverse=True):
+ fixedv = k.split(",")[0]
+
+ sectioncves = []
+ for cve in sorted(v, key=lambda s: [int(u) if u.isdigit() else u for u in s["id"].split('-')]):
+ e = {}
+ e['cveid'] = cve["id"]
+ data_version = cve.get("dataVersion", DEFAULT_CVE_DATA_VERSION)
+ if data_version.startswith("5"):
+ try:
+ e['impact'] = cve["containers"]["cna"]["metrics"][0]["other"]["content"]["text"]
+ e['title'] = cve["containers"]["cna"]["title"]
+ e['desc'] = cve["containers"]["cna"]["descriptions"][0]["value"]
+ e['credit'] = []
+ if ("credits" in cve["containers"]["cna"]):
+ for credit in cve["containers"]["cna"]["credits"]:
+ e['credit'].append(credit["type"]+": "+credit["value"])
+ except Exception as err:
+ sys.stderr.write("Missing data in " + cve["_filename"] + ": " + traceback.format_exc().replace('\n','\n '))
+ sys.exit(1);
+ affects = []
+ product = cve["containers"]["cna"]["affected"][0]
+ productname = product['product']
+ for ver in product["versions"]:
+ base = ""
+ appended = False
+ if ver.get("version", "0") != "0":
+ base = ver["version"]
+ if "lessThanOrEqual" in ver:
+ affects.append(f"{base} through {ver['lessThanOrEqual']}")
+ appended = True
+ if "lessThan" in ver:
+ affects.append(f"{base} before {ver['lessThan']}")
+ appended = True
+ if not appended and base != "":
+ affects.append(base)
+ # Make a natural order sort
+ affects.sort(reverse=True, key=natural_sort_key)
+ e['affects'] = ", ".join(affects)
+ e['timetable'] = [];
+ for time in cve["containers"]["cna"].get("timeline", []): # Timeline may not be present in the JSON!
+ timed = time["value"]
+ if ("reported" in timed):
+ timed = "Reported to security team"
+ elif ("public" in timed):
+ timed = "Issue public"
+ elif ("release" in timed):
+ timed = "Update "+timed
+ e['timetable'].append([timed,time["time"].split('T')[0]])
+ sectioncves.append(e)
+ else:
+ e['impact'] = cve["impact"][0]["other"]
+ e['title'] = cve["CVE_data_meta"]["TITLE"]
+ e['desc'] = cve["description"]["description_data"][0]["value"]
+ e['credit'] = []
+ if ("credit" in cve):
+ for credit in cve["credit"]:
+ e['credit'].append(credit["value"])
+ affects = []
+ product = cve["affects"]["vendor"]["vendor_data"][0]["product"]["product_data"][0]
+ productname = product['product_name']
+ for ver in product["version"]["version_data"]:
+ if (ver["version_affected"] == "="):
+ affects.append(ver["version_value"])
+ elif (ver["version_affected"] == "?="):
+ # We did ?= for "maybe affects" because no one checked
+ affects.append(ver["version_value"]+"?")
+ else:
+ # Otherwise maybe we started doing things like "<2.7.8"
+ affects.append(ver["version_affected"]+ver["version_value"])
+ # Make a natural order sort
+ affects.sort(reverse=True, key=natural_sort_key)
+ e['affects'] = ", ".join(affects)
+ e['timetable'] = [];
+ for time in cve.get("timeline", []):
+ timed = time["value"]
+ if ("reported" in timed):
+ timed = "Reported to security team"
+ elif ("public" in timed):
+ timed = "Issue public"
+ elif ("release" in timed):
+ timed = "Update "+timed
+ e['timetable'].append([timed,time["time"]])
+ sectioncves.append(e)
+ sections.append({"cves":sectioncves,"fixed":fixedv,"product":productname})
+
+# Everything is sorted and pretty, this should be some python template thing
+
+# We are generating html in markdown. Add the metadata first.
+if not options.nopreamble:
+ print ("Title: "+productname+" "+filterversion+" vulnerabilities")
+ print ("asf_headings: False")
+ print ("")
+ print ("<h1>"+productname+" "+filterversion+" vulnerabilities</h1>")
+ print ("<p>This page lists all security vulnerabilities fixed in released versions of "+productname+" "+filterversion+". Each vulnerability is given a security <a href=\""+options.impactlink+"\">impact rating</a> by the Apache security team - please note that this rating may well vary from platform to platform. We also list the versions the flaw is known to affect, and where a flaw has not been verified list the version with a question mark.</p>")
+ print ("<p>Please note that if a vulnerability is shown below as being fixed in a \"-dev\" release then this means that a fix has been applied to the development source tree and will be part of an upcoming full release.</p>")
+ print ("<p>Please send comments or corrections for these vulnerabilities to the <a href=\"/security_report.html\">Security Team</a>.</p> <br/>")
+
+ if (options.extratext):
+ print ("<p>"+options.extratext+"</p><br/>")
+
+for sectioncves in sections:
+ print ("\n<h1 id=\""+sectioncves["fixed"]+"\">Fixed in "+sectioncves["product"]+" "+sectioncves["fixed"]+"</h1><dl>\n")
+ for e in sectioncves["cves"]:
+ html = "<dt><h3 id=\""+e['cveid']+"\">"+e['impact']+": <name name=\""+e['cveid']+"\">"+saxutils.escape(e['title'])+"</name>\n";
+ html += "(<a href=\"https://www.cve.org/CVERecord?id="+e['cveid']+"\">"+e['cveid']+"</a>)</h3></dt>\n";
+ desc = saxutils.escape(e['desc'])
+ desc = re.sub(r'\n','</p><p>', desc)
+ html += "<dd><p>"+desc+"</p>\n"
+ if (e['credit']):
+ html += "<p>Acknowledgements:"
+ if len(e['credit']) == 1:
+ html += " " + saxutils.escape(e['credit'][0]) + "</p>\n"
+ else:
+ html += "</p>\n<ul>\n"
+ for credit in e['credit']:
+ html += "<li>" + saxutils.escape(credit) + "</li>\n"
+ html += "</ul>\n"
+ html += "<table class=\"cve\">"
+ e['timetable'].append(["Affects",e['affects']]);
+ for ti in e['timetable']:
+ html+= "<tr><td class=\"cve-header\">"+ti[0]+"</td><td class=\"cve-value\">"+ti[1]+"</td></tr>\n"
+ html+= "</table></dd>"
+ print (html)
+ print ("</dl></br/>")
+
+