Reconcile README and ingest command implementation

diff --git a/README.rst b/README.rst
index 97f4ecf39a58d48a0901df26c4dded3e82849275..ebacacf2397e0e15308f9558ee207e64632bad15 100644 (file)
--- a/README.rst
+++ b/README.rst
@@ -237,17 +237,22 @@ presence elsewhere in the database.  We can enumerate files not tracked using
 above::
 
    find ${DIR} -type f -print0 | \
-     cdb --db ${DB} filterpath --in-path --predicate=out -0 -1 --format '$u$z' > ${DB}.new-files0
-   xargs -0 sha512sum > ${DB}.new < ${DB}.new-files0
+   cdb --db ${DB} filterpath --in-path --predicate=out -0 -1 --format '$u$z' | \
+   xargs -0 sha512sum > ${DB}.new
 
 We can then prepare to prune duplicates and add unique files::
 
-   cdb --db ${DB} ingest -1 --prune --inplace --digest-log ${DB}.new2 < ${DB}.new > ${DB}.prune
+   cdb --db ${DB} ingest --prune --inplace --digest-log ${DB}.new2 < ${DB}.new > ${DB}.prune
 
 Add new files to the database with::
 
    cdb --db ${DB} addh < ${DB}.new2
 
-After reviewing the files to be pruned in ``${DB}.prune``, it can be executed::
+Inspect the pruning commands to be run, and then execute them with::
 
    sh < ${DB}.prune
+
+(If you have, or might have, unusual path names, you may be better served with
+``--prune-log`` rather than ``--prune``.  The resulting, ``NUL``-terminated list
+of files can be inspected with ``cdb-util escape human -0`` and run with ``xargs
+-0 -- rm --``.)

diff --git a/cdb b/cdb
index 418ef5b99d9dc8eb88bd8de7f890aa36863761c6..d277f20b5befe29527aab565d1a31cd19c185ebf 100755 (executable)
--- a/cdb
+++ b/cdb
@@ -223,9 +223,10 @@ mksubcmd(function(c)
    :description("Don't crush targets to basename; trim given prefix instead")
   local mv = c:flag("--move")
    :description("Issue move, not copy, commands for new files")
-  c:option("--prune-log")
-   :argname("file")
-   :description("Log of files with colliding digests; NUL-separated")
+  c:option("--prune")
+   :argname("LOG")
+   :default(""):defmode("a")
+   :description("Generate rm for, or log w/ NULs, files with colliding digests")
   c:option("--digest-log")
    :argname("file")
    :description("Write novel digest records to the indicated file")
@@ -256,11 +257,21 @@ mksubcmd(function(c)
     end
   end
 
+  local shell_escape =
+    args.extended_escapes and cdblib.extended_shell_escape
+                          or  cdblib.posix_shell_escape
+
   local function log_prune() end
-  if args.prune_log then
-    local prunef = assert(io.open(args.prune_log, "wb"))
-    function log_prune(p)
-      return prunef:write(p, "\0")
+  if args.prune then
+    if args.prune == "" then
+      function log_prune(p)
+        return io.write("rm -- ", shell_escape(p), eol)
+      end
+    else
+      local prunef = assert(io.open(args.prune, "wb"))
+      function log_prune(p)
+        return prunef:write(p, "\0")
+      end
     end
   end
 
@@ -270,9 +281,6 @@ mksubcmd(function(c)
   local sth_path_by_hash = assert(sql_mk_pathid_find_by_hash(dbh))
   local sth_superseder_by_hash = assert(sql_mk_superseder_find_by_hash(dbh))
 
-  local shell_escape =
-    args.extended_escapes and cdblib.extended_shell_escape
-                          or  cdblib.posix_shell_escape
   local human_escape = cdblib.extended_shell_escape
   if not args.no_human_escape then
     local ok, rexlib = pcall(require, "rex_posix")

diff --git a/test.sh b/test.sh
index 9e087d4e060c18cfcbb3886b079e6c8b7ca6db39..f7113513012998457d3d1a1cee6ef4cc2ed30049 100755 (executable)
--- a/test.sh
+++ b/test.sh
@@ -131,12 +131,12 @@ diff -u - <(${LUA} ./cdb --db ${DB1} mappath <<<"ordinary") <<<"1  ordinary"
 # Test 'ingeset'
 cat >${LOG1} <<HERE
 5  new
-\\6  new\\\\esc
 9  twinned copy 🎵
+\\6  new\\\\esc
 4  ti'cky copy
 4  ti'cky copy with \$extra
 HERE
-${LUA} ./cdb --db ${DB1} ingest --target x --prune-log=${LOG4} --verbose <${LOG1} >${LOG2} 2>${LOG3}
+${LUA} ./cdb --db ${DB1} ingest --target x --prune=${LOG4} --verbose <${LOG1} >${LOG2} 2>${LOG3}
 # Import commands on stdout
 diff -u - ${LOG2} <<HERE
 cp -- 'new' 'x/new'
@@ -145,8 +145,8 @@ HERE
 # Log on stderr
 diff -u - ${LOG3} <<HERE
 Import 'new' to 'x/new'
-Import 'new\\esc' to 'x/new\\esc'
 Import hash 9 from path 'twinned copy 🎵' already in database at 'twinned'
+Import 'new\\esc' to 'x/new\\esc'
 Import hash 4 from path "ti'cky copy" already in database at "ti'cky"
 Import hash 4 from path 'ti'"'"'cky copy with \$extra' already in database at "ti'cky"
 HERE
@@ -157,6 +157,18 @@ ti'cky copy
 ti'cky copy with \$extra
 HERE
 
+# Again, with intermixed pruning commands
+${LUA} ./cdb --db ${DB1} ingest --target x --prune <${LOG1} >${LOG2} 2>${LOG3}
+diff -u - ${LOG2} <<HERE
+cp -- 'new' 'x/new'
+rm -- 'twinned copy 🎵'
+cp -- 'new\\esc' 'x/new\\esc'
+rm -- 'ti'"'"'cky copy'
+rm -- 'ti'"'"'cky copy with \$extra'
+HERE
+diff -u /dev/null ${LOG3}
+
+# With move
 diff -u - <(${LUA} ./cdb --db ${DB1} ingest --move --target x --digest-log ${LOG2} <${LOG1}) <<HERE
 mv -- 'new' 'x/new'
 mv -- 'new\\esc' 'x/new\\esc'
@@ -169,7 +181,7 @@ HERE
 diff -u /dev/null <(${LUA} ./cdb --db ${DB1} gc)
 
 # And without actually doing the import
-${LUA} ./cdb --db ${DB1} ingest --prune-log=${LOG3} --verbose >${LOG1} 2>${LOG2} \
+${LUA} ./cdb --db ${DB1} ingest --prune=${LOG3} --verbose >${LOG1} 2>${LOG2} \
   <<<'5  new'$'\n''4  copy'
 # No output stdout
 diff -u /dev/null ${LOG1}
@@ -291,7 +303,7 @@ diff -u /dev/null <(${LUA} ./cdb --db ${DB1} maph 11)
 diff -u /dev/null <(${LUA} ./cdb --db ${DB1} gc)
 
 # Test ingest with supers
-${LUA} ./cdb --db ${DB1} ingest --target x --prune-log=${LOG3} --digest-log=${LOG4} --verbose \
+${LUA} ./cdb --db ${DB1} ingest --target x --prune=${LOG3} --digest-log=${LOG4} --verbose \
   >${LOG1} 2>${LOG2} <<HERE
 0  won't fix super
 11  pre-ordinary again