From 5fa4379088f05e149799780615ebf8a2bea91a00 Mon Sep 17 00:00:00 2001
From: B Stack <bgstack15@gmail.com>
Date: Thu, 27 Feb 2020 08:25:22 -0500
Subject: refactor again

   parse Packages.gz.mirrorlist, and choose a single mirror to pull
   all contents from.
---
 obsmirror.sh/obsmirror.sh | 20 +++++++++++++++-----
 1 file changed, 15 insertions(+), 5 deletions(-)

diff --git a/obsmirror.sh/obsmirror.sh b/obsmirror.sh/obsmirror.sh
index 655bf2f..86180f3 100755
--- a/obsmirror.sh/obsmirror.sh
+++ b/obsmirror.sh/obsmirror.sh
@@ -42,7 +42,6 @@ parse_obs_page_and_subdirs() {
          curl -s -L "${___inurl}/${subdir}" | parse_obs_dl_page | sed -r -e "s@^@${___inurl}/${subdir}@"
       done
    } > "${___tmpfile}"
-
 }
 
 test -n "${OBSMIRROR_CONF}" && . "${OBSMIRROR_CONF}"
@@ -57,11 +56,22 @@ test -z "${thisuser}" && thisuser=obsmirror
    test "${DEBUG:-NONE}" = "FULL" && set -x
       echo "logfile=${logfile}"
 
-   mkdir -p "${workdir}" ; chmod "0711" "${workdir}" ; chown "${thisuser}:$( id -G "${thisuser}" | awk '{print $1}' )" "${workdir}"
+   mkdir -p "${workdir}" ; chmod "0755" "${workdir}" ; chown "${thisuser}:$( id -G "${thisuser}" | awk '{print $1}' )" "${workdir}"
    cd "${workdir}"
-   # get page contents
+   # get mirrorlist of Packages.gz file and find the one that lists the most packages
+   step1="$( curl -s -L "${inurl}/Packages.gz.mirrorlist" )"
+   options="$( echo "${step1}" | grep -oE 'href="[^"]+">' | awk '!x[$0]++' | sed -r -e 's/^href="//;' -e 's/">$//;' | grep -iE '^(ht|f)tps?:\/\/.*Packages\.gz$' )"
+   results="$(
+   for entry in ${options} ; do
+      curl -s -L "${entry}" | zgrep -cE '^Package:' | sed -r -e "s@\$@ ${entry}@;"
+   done )"
+   topresult_line="$( echo "${results}" | sort -nr | head -n1 | sed -r -e 's/\/Packages\.gz$//;' )"
+   topresult_packagecount="$( echo "${topresult_line}" | awk '{print $1}' )"
+   topresult="$( echo "${topresult_line}" | awk '{print $2}' )"
+   echo "USING ${topresult} with ${topresult_packagecount} packages" 1>&2
+   inurl="${topresult}"
+
    step1="$( curl -s -L "${inurl}" )"
-   #step2="$( echo "${step1}" | parse_obs_dl_page )"
    echo "${step1}" | parse_obs_page_and_subdirs "${inurl}" "${tmpfile}"
    # loop over all entries and download them
    for thisurl in $( cat "${tmpfile}" ) ;
@@ -70,7 +80,7 @@ test -z "${thisuser}" && thisuser=obsmirror
       thisdir="$( dirname "${thisfile}" )"
       test -d "${thisdir}" || mkdir -p "${thisdir}"
       test -n "${VERBOSE}" && echo "FROM ${thisurl} TO ${thisfile}"
-      test -z "${DRYRUN}" && wget --continue --no-verbose -O "${thisfile}" "${thisurl}" &
+      test -z "${DRYRUN}" && wget --no-verbose -O "${thisfile}" "${thisurl}" &
    done
 
 } 2>&1 | tee -a "${logfile}"
-- 
cgit