FIX: Adapt deduplicate sed command for OS X.

The sed command of BSD / OS X doesn't like { command } blocks in a single expression. So move the (dynamic) deduplicate sed command(s) to the end, and use a label to access them. This also makes the entire sed script somewhat easier to understand.

README.textile

@@ -21,6 +21,7 @@ h2. Quick Links
 * Original anemic release by "Gina Trapani":http://ginatrapani.org on 5/11/2006.
 * Raised to great heights by "brainy and dedicated volunteers":http://github.com/ginatrapani/todo.txt-cli/network.
 * Licensed under the "GPL":http://www.gnu.org/copyleft/gpl.html
-* "Add-on Directory":http://wiki.github.com/ginatrapani/todo.txt-cli/todosh-add-on-directory
+* "Add-on Directory":https://github.com/ginatrapani/todo.txt-cli/wiki/Todo.sh-Add-on-Directory
+https://github.com/ginatrapani/todo.txt-cli/wiki/Creating-and-Installing-Add-ons
 * "Changelog":http://wiki.github.com/ginatrapani/todo.txt-cli/todosh-changelog
-* "Known Bugs":http://github.com/ginatrapani/todo.txt-cli/issues
+* "Known Bugs":http://github.com/ginatrapani/todo.txt-cli/issues

tests/t1300-ls.sh

@@ -53,6 +53,18 @@ test_todo_session 'checking TODOTXT_FINAL_FILTER' <<EOF
 TODO: 3 of 3 tasks shown
 EOF
 
+#
+# check the custom hiding
+#
+test_todo_session 'checking HIDE_CUSTOM_SUBSTITUTION' <<EOF
+>>> HIDE_CUSTOM_SUBSTITUTION='[tT]h' todo.sh ls
+2 aaa zzz is line should be first.
+3 bbb yyy is line should be second.
+1 ccc xxx is line should be ird.
+--
+TODO: 3 of 3 tasks shown
+EOF
+
 #
 # check the filtering of TERM
 #

tests/t1900-archive.sh

@@ -0,0 +1,35 @@
+#!/bin/sh
+
+test_description='archive functionality
+
+Ensure we can archive items successfully.
+'
+. ./test-lib.sh
+
+cat > todo.txt <<EOF
+one
+two
+three
+one
+x done
+four
+EOF
+
+test_todo_session 'archive with duplicates' <<EOF
+>>> todo.sh archive
+x done
+TODO: $HOME/todo.txt archived.
+EOF
+
+test_todo_session 'list after archive' <<EOF
+>>> todo.sh ls
+5 four
+1 one
+4 one
+3 three
+2 two
+--
+TODO: 5 of 5 tasks shown
+EOF
+
+test_done

tests/t1910-deduplicate.sh

@@ -0,0 +1,103 @@
+#!/bin/sh
+
+test_description='deduplicate functionality
+
+Ensure we can deduplicate items successfully.
+'
+. ./test-lib.sh
+
+cat > todo.txt <<EOF
+duplicated
+two
+x done
+duplicated
+double task
+double task
+three
+EOF
+
+test_todo_session 'deduplicate and preserve line numbers' <<EOF
+>>> todo.sh deduplicate
+TODO: 2 duplicate task(s) removed
+
+>>> todo.sh -p ls
+5 double task
+1 duplicated
+7 three
+2 two
+3 x done
+--
+TODO: 5 of 5 tasks shown
+EOF
+
+test_todo_session 'deduplicate without duplicates' <<EOF
+>>> todo.sh deduplicate
+TODO: No duplicate tasks found
+EOF
+
+cat > todo.txt <<EOF
+duplicated
+two
+x done
+duplicated
+double task
+double task
+three
+EOF
+test_todo_session 'deduplicate and delete lines' <<EOF
+>>> todo.sh -n deduplicate
+TODO: 2 duplicate task(s) removed
+
+>>> todo.sh -p ls
+4 double task
+1 duplicated
+5 three
+2 two
+3 x done
+--
+TODO: 5 of 5 tasks shown
+EOF
+
+cat > todo.txt <<EOF
+one
+duplicated
+three
+duplicated
+duplicated
+six
+duplicated
+EOF
+test_todo_session 'deduplicate more than two occurrences' <<EOF
+>>> todo.sh deduplicate
+TODO: 3 duplicate task(s) removed
+
+>>> todo.sh -p ls
+2 duplicated
+1 one
+6 six
+3 three
+--
+TODO: 4 of 4 tasks shown
+EOF
+
+cat > todo.txt <<EOF
+normal task
+a bold task
+something else
+a bold task
+something more
+EOF
+test_todo_session 'deduplicate with non-printable duplicates' <<EOF
+>>> todo.sh deduplicate
+TODO: 1 duplicate task(s) removed
+
+>>> todo.sh -p ls
+2 a bold task
+1 normal task
+3 something else
+5 something more
+--
+TODO: 4 of 4 tasks shown
+EOF
+
+test_done

tests/t1950-report.sh

@@ -0,0 +1,96 @@
+#!/bin/sh
+
+test_description='report functionality
+
+This test checks the reporting and the format of the report file.
+'
+. ./test-lib.sh
+
+cat > todo.txt <<EOF
+(B) smell the uppercase Roses +flowers @outside
+stop and think
+smell the coffee +wakeup
+make the coffee +wakeup
+visit http://example.com
+EOF
+
+test_todo_session 'create new report' <<EOF
+>>> todo.sh report
+TODO: $HOME/todo.txt archived.
+2009-02-13T04:40:00 5 0
+TODO: Report file updated.
+
+>>> todo.sh -p list
+1 (B) smell the uppercase Roses +flowers @outside
+4 make the coffee +wakeup
+3 smell the coffee +wakeup
+2 stop and think
+5 visit http://example.com
+--
+TODO: 5 of 5 tasks shown
+EOF
+
+test_todo_session 'report of done tasks' <<EOF
+>>> todo.sh -A do 3
+3 x 2009-02-13 smell the coffee +wakeup
+TODO: 3 marked as done.
+x 2009-02-13 smell the coffee +wakeup
+TODO: $HOME/todo.txt archived.
+
+>>> todo.sh report
+TODO: $HOME/todo.txt archived.
+2009-02-13T04:40:00 4 1
+TODO: Report file updated.
+
+>>> todo.sh -p list
+1 (B) smell the uppercase Roses +flowers @outside
+3 make the coffee +wakeup
+2 stop and think
+4 visit http://example.com
+--
+TODO: 4 of 4 tasks shown
+EOF
+
+test_todo_session 'report performs archiving' <<EOF
+>>> todo.sh -a do 3
+3 x 2009-02-13 make the coffee +wakeup
+TODO: 3 marked as done.
+
+>>> todo.sh report
+x 2009-02-13 make the coffee +wakeup
+TODO: $HOME/todo.txt archived.
+2009-02-13T04:40:00 3 2
+TODO: Report file updated.
+
+>>> todo.sh -p list
+1 (B) smell the uppercase Roses +flowers @outside
+2 stop and think
+3 visit http://example.com
+--
+TODO: 3 of 3 tasks shown
+
+>>> todo.sh -p listfile done.txt
+2 x 2009-02-13 make the coffee +wakeup
+1 x 2009-02-13 smell the coffee +wakeup
+--
+DONE: 2 of 2 tasks shown
+EOF
+
+test_todo_session 'report is unchanged when no changes' <<EOF
+>>> cat report.txt
+2009-02-13T04:40:00 5 0
+2009-02-13T04:40:00 4 1
+2009-02-13T04:40:00 3 2
+
+>>> todo.sh report
+TODO: $HOME/todo.txt archived.
+2009-02-13T04:40:00 3 2
+TODO: Report file is up-to-date.
+
+>>> cat report.txt
+2009-02-13T04:40:00 5 0
+2009-02-13T04:40:00 4 1
+2009-02-13T04:40:00 3 2
+EOF
+
+test_done

tests/t9999-testsuite_example.sh

@@ -85,13 +85,9 @@ TODO: $HOME/todo.txt archived.
 TODO: 5 of 5 tasks shown
 
 >>> todo.sh report
+TODO: $HOME/todo.txt archived.
+2009-02-13T04:40:00 5 1
 TODO: Report file updated.
-2009-02-13-04:40:00 5 1
-
->>> todo.sh report
-TODO: Report file updated.
-2009-02-13-04:40:00 5 1
-2009-02-13-04:40:00 5 1
 
 >>> todo.sh append g a
 usage: todo.sh append ITEM# "TEXT TO APPEND"

todo.sh

@@ -50,6 +50,7 @@ shorthelp()
 		    append|app ITEM# "TEXT TO APPEND"
 		    archive
 		    command [ACTIONS]
+		    deduplicate
 		    del|rm ITEM# [TERM]
 		    depri|dp ITEM#[, ITEM#, ITEM#, ...]
 		    do ITEM#[, ITEM#, ITEM#, ...]
@@ -67,6 +68,8 @@ shorthelp()
 		    report
 		    shorthelp
 
+		  Actions can be added and overridden using scripts in the actions
+		  directory.
 	EndHelp
 
     # Only list the one-line usage from the add-on actions. This assumes that
@@ -87,10 +90,10 @@ help()
 
 		  Options:
 		    -@
-		        Hide context names in list output. Use twice to show context
+		        Hide context names in list output.  Use twice to show context
 		        names (default).
 		    -+
-		        Hide project names in list output. Use twice to show project
+		        Hide project names in list output.  Use twice to show project
 		        names (default).
 		    -c
 		        Color mode
@@ -103,7 +106,7 @@ help()
 		    -p
 		        Plain mode turns off colors
 		    -P
-		        Hide priority labels in list output. Use twice to show
+		        Hide priority labels in list output.  Use twice to show
 		        priority labels (default).
 		    -a
 		        Don't auto-archive tasks automatically on completion
@@ -162,7 +165,6 @@ help()
 		      Adds FIRST THING I NEED TO DO to your todo.txt on its own line and
 		      Adds SECOND THING I NEED TO DO to you todo.txt on its own line.
 		      Project and context notation optional.
-		      Quotes optional.
 
 		    addto DEST "TEXT TO ADD"
 		      Adds a line of text to any file located in the todo.txt directory.
@@ -180,6 +182,9 @@ help()
 		      Runs the remaining arguments using only todo.sh builtins.
 		      Will not call any .todo.actions.d scripts.
 
+		    deduplicate
+		      Removes duplicate lines from todo.txt.
+
 		    del ITEM# [TERM]
 		    rm ITEM# [TERM]
 		      Deletes the task on line ITEM# in todo.txt.
@@ -199,13 +204,19 @@ help()
 		    list [TERM...]
 		    ls [TERM...]
 		      Displays all tasks that contain TERM(s) sorted by priority with line
-		      numbers.  If no TERM specified, lists entire todo.txt.
+		      numbers.  Each task must match all TERM(s) (logical AND); to display
+		      tasks that contain any TERM (logical OR), use
+		      "TERM1\|TERM2\|..." (with quotes), or TERM1\\\|TERM2 (unquoted).
+		      Hides all tasks that contain TERM(s) preceded by a
+		      minus sign (i.e. -TERM). If no TERM specified, lists entire todo.txt.
 
 		    listall [TERM...]
 		    lsa [TERM...]
 		      Displays all the lines in todo.txt AND done.txt that contain TERM(s)
-		      sorted by priority with line  numbers.  If no TERM specified, lists
-		      entire todo.txt AND done.txt concatenated and sorted.
+		      sorted by priority with line  numbers.  Hides all tasks that
+		      contain TERM(s) preceded by a minus sign (i.e. -TERM).  If no
+		      TERM specified, lists entire todo.txt AND done.txt
+		      concatenated and sorted.
 
 		    listcon
 		    lsc
@@ -215,19 +226,23 @@ help()
 		    lf [SRC [TERM...]]
 		      Displays all the lines in SRC file located in the todo.txt directory,
 		      sorted by priority with line  numbers.  If TERM specified, lists
-		      all lines that contain TERM in SRC file.
+		      all lines that contain TERM(s) in SRC file.  Hides all tasks that
+		      contain TERM(s) preceded by a minus sign (i.e. -TERM).  
 		      Without any arguments, the names of all text files in the todo.txt
 		      directory are listed.
-
+		
 		    listpri [PRIORITY] [TERM...]
 		    lsp [PRIORITY] [TERM...]
 		      Displays all tasks prioritized PRIORITY.
 		      If no PRIORITY specified, lists all prioritized tasks.
-		      If TERM specified, lists only prioritized tasks that contain TERM.
+		      If TERM specified, lists only prioritized tasks that contain TERM(s).
+		      Hides all tasks that contain TERM(s) preceded by a minus sign
+		      (i.e. -TERM).  
 
 		    listproj
 		    lsprj
-		      Lists all the projects that start with the + sign in todo.txt.
+		      Lists all the projects (terms that start with a + sign) in
+		      todo.txt.
 
 		    move ITEM# DEST [SRC]
 		    mv ITEM# DEST [SRC]
@@ -245,7 +260,7 @@ help()
 		    p ITEM# PRIORITY
 		      Adds PRIORITY to task on line ITEM#.  If the task is already
 		      prioritized, replaces current priority with new PRIORITY.
-		      PRIORITY must be an uppercase letter between A and Z.
+		      PRIORITY must be a letter between A and Z.
 
 		    replace ITEM# "UPDATED TODO"
 		      Replaces task on line ITEM# with UPDATED TODO.
@@ -256,7 +271,6 @@ help()
 		    shorthelp
 		      List the one-line usage of all built-in and add-on actions.
 
-
 	EndActionsHelp
 
         addonHelp
@@ -353,20 +367,6 @@ getNewtodo()
     [ -z "$newtodo" ] && die "$(getPrefix "$2"): No updated task $item."
 }
 
-archive()
-{
-    #defragment blank lines
-    sed -i.bak -e '/./!d' "$TODO_FILE"
-    [ $TODOTXT_VERBOSE -gt 0 ] && grep "^x " "$TODO_FILE"
-    grep "^x " "$TODO_FILE" >> "$DONE_FILE"
-    sed -i.bak '/^x /d' "$TODO_FILE"
-    cp "$TODO_FILE" "$TMP_FILE"
-    sed -n 'G; s/\n/&&/; /^\([ ~-]*\n\).*\n\1/d; s/\n//; h; P' "$TMP_FILE" > "$TODO_FILE"
-    if [ $TODOTXT_VERBOSE -gt 0 ]; then
-	echo "TODO: $TODO_FILE archived."
-    fi
-}
-
 replaceOrPrepend()
 {
   action=$1; shift
@@ -775,7 +775,11 @@ _list() {
     fi
     items=$(
         sed = "$src"                                            \
-        | sed "N; s/^/     /; s/ *\(.\{$PADDING,\}\)\n/\1 /"    \
+        | sed '''
+            N
+            s/^/     /
+            s/ *\([ 0-9]\{'"$PADDING"',\}\)\n/\1 /
+	  '''                                                 \
         | grep -v "^[ 0-9]\+ *$"
     )
     if [ "${filter_command}" ]; then
@@ -812,8 +816,9 @@ _list() {
             }
           '''  \
         | sed '''
-            s/'${HIDE_PROJECTS_SUBSTITUTION:-^}'//g
-            s/'${HIDE_CONTEXTS_SUBSTITUTION:-^}'//g
+            s/'"${HIDE_PROJECTS_SUBSTITUTION:-^}"'//g
+            s/'"${HIDE_CONTEXTS_SUBSTITUTION:-^}"'//g
+            s/'"${HIDE_CUSTOM_SUBSTITUTION:-^}"'//g
           '''                                                   \
         | eval ${TODOTXT_FINAL_FILTER}                          \
     )
@@ -933,7 +938,15 @@ case $action in
     ;;
 
 "archive" )
-    archive;;
+    # defragment blank lines
+    sed -i.bak -e '/./!d' "$TODO_FILE"
+    [ $TODOTXT_VERBOSE -gt 0 ] && grep "^x " "$TODO_FILE"
+    grep "^x " "$TODO_FILE" >> "$DONE_FILE"
+    sed -i.bak '/^x /d' "$TODO_FILE"
+    if [ $TODOTXT_VERBOSE -gt 0 ]; then
+	echo "TODO: $TODO_FILE archived."
+    fi
+    ;;
 
 "del" | "rm" )
     # replace deleted line with a blank line when TODOTXT_PRESERVE_LINE_NUMBERS is 1
@@ -991,7 +1004,7 @@ case $action in
 
     # Split multiple depri's, if comma separated change to whitespace separated
     # Loop the 'depri' function for each item
-    for item in $(echo $* | tr ',' ' '); do
+    for item in ${*//,/ }; do
         getTodo "$item"
 
 	if [[ "$todo" = \(?\)\ * ]]; then
@@ -1015,7 +1028,7 @@ case $action in
 
     # Split multiple do's, if comma separated change to whitespace separated
     # Loop the 'do' function for each item
-    for item in $(echo $* | tr ',' ' '); do
+    for item in ${*//,/ }; do
         getTodo "$item"
 
         # Check if this item has already been done
@@ -1035,7 +1048,9 @@ case $action in
     done
 
     if [ $TODOTXT_AUTO_ARCHIVE = 1 ]; then
-        archive
+        # Recursively invoke the script to allow overriding of the archive
+        # action.
+        "$TODO_FULL_SH" archive
     fi
     ;;
 
@@ -1200,22 +1215,69 @@ note: PRIORITY must be anywhere from A to Z."
     ;;
 
 "report" )
-    #archive first
-    sed '/^x /!d' "$TODO_FILE" >> "$DONE_FILE"
-    sed -i.bak '/^x /d' "$TODO_FILE"
+    # archive first
+    # Recursively invoke the script to allow overriding of the archive
+    # action.
+    "$TODO_FULL_SH" archive
 
-    NUMLINES=$( sed -n '$ =' "$TODO_FILE" )
-    if [ ${NUMLINES:-0} = "0" ]; then
-         echo "datetime todos dones" >> "$REPORT_FILE"
-    fi
-    #now report
     TOTAL=$( sed -n '$ =' "$TODO_FILE" )
     TDONE=$( sed -n '$ =' "$DONE_FILE" )
-    TECHO=$(echo $(date +%Y-%m-%d-%T); echo ' '; echo ${TOTAL:-0}; echo ' ';
-    echo ${TDONE:-0})
-    echo $TECHO >> "$REPORT_FILE"
-    [ $TODOTXT_VERBOSE -gt 0 ] && echo "TODO: Report file updated."
-    cat "$REPORT_FILE"
+    NEWDATA="${TOTAL:-0} ${TDONE:-0}"
+    LASTREPORT=$(sed -ne '$p' "$REPORT_FILE")
+    LASTDATA=${LASTREPORT#* }   # Strip timestamp.
+    if [ "$LASTDATA" = "$NEWDATA" ]; then
+        echo "$LASTREPORT"
+        [ $TODOTXT_VERBOSE -gt 0 ] && echo "TODO: Report file is up-to-date."
+    else
+        NEWREPORT="$(date +%Y-%m-%dT%T) ${NEWDATA}"
+        echo "${NEWREPORT}" >> "$REPORT_FILE"
+        echo "${NEWREPORT}"
+        [ $TODOTXT_VERBOSE -gt 0 ] && echo "TODO: Report file updated."
+    fi
+    ;;
+
+"deduplicate" )
+    if [ $TODOTXT_PRESERVE_LINE_NUMBERS = 0 ]; then
+        deduplicateSedCommand='d'
+    else
+        deduplicateSedCommand='s/^.*//; p'
+    fi
+
+    # To determine the difference when deduplicated lines are preserved, only
+    # non-empty lines must be counted.
+    originalTaskNum=$( sed -e '/./!d' "$TODO_FILE" | sed -n '$ =' )
+
+    # Look for duplicate lines and discard the second occurrence.
+    # We start with an empty hold space on the first line.  For each line:
+    #   G - appends newline + hold space to the pattern space
+    #   s/\n/&&/; - double up the first new line so we catch adjacent dups
+    #   /^\([^\n]*\n\).*\n\1/b dedup
+    #       If the first line of the hold space shows up again later as an
+    #       entire line, it's a duplicate. Jump to the "dedup" label, where
+    #       either of the following is executed, depending on whether empty
+    #       lines should be preserved:
+    #       d           - Delete the current pattern space, quit this line and
+    #                     move on to the next, or:
+    #       s/^.*//; p  - Clear the task text, print this line and move on to
+    #                     the next.
+    #   s/\n//;   - else (no duplicate), drop the doubled newline
+    #   h;        - replace the hold space with the expanded pattern space
+    #   P;        - print up to the first newline (that is, the input line)
+    #   b         - end processing of the current line
+    sed -i.bak -n \
+        -e 'G; s/\n/&&/; /^\([^\n]*\n\).*\n\1/b dedup' \
+        -e 's/\n//; h; P; b' \
+        -e ':dedup' \
+        -e "$deduplicateSedCommand" \
+        "$TODO_FILE"
+
+    newTaskNum=$( sed -e '/./!d' "$TODO_FILE" | sed -n '$ =' )
+    deduplicateNum=$(( originalTaskNum - newTaskNum ))
+    if [ $deduplicateNum -eq 0 ]; then
+        echo "TODO: No duplicate tasks found"
+    else
+        echo "TODO: $deduplicateNum duplicate task(s) removed"
+    fi
     ;;
 
 * )

todo_completion

@@ -38,20 +38,28 @@ _todo()
                     +*) completions=$(TODOTXT_VERBOSE=0 todo.sh command listproj);;
                     @*) completions=$(TODOTXT_VERBOSE=0 todo.sh command listcon);;
                     *)  if [[ "$cur" =~ ^[0-9]+$ ]]; then
-                            local item=$(TODOTXT_VERBOSE=0 todo.sh -@ -+ -p -x command ls "^ *${cur} " | head -n 1)
-
                             # Remove the (padded) task number; we prepend the
-                            # user-provided $cur.
-                            item=${item#* }
-
-                            # Remove the timestamp prepended by the -t option;
-                            # there's no todo.txt option for that yet.
-                            item=${item#[0-9][0-9][0-9][0-9]-[0-9][0-9]-[0-9][0-9] }
-
+                            # user-provided $cur instead.
+                            # Remove the timestamp prepended by the -t option,
+                            # and the done date (for done tasks); there's no
+                            # todo.txt option for that yet.
+                            # But keep priority and "x"; they're short and may
+                            # provide useful context.
+                            # Remove any trailing whitespace; the Bash
+                            # completion inserts a trailing space itself.
+                            # Finally, limit the output to a single line just as
+                            # a safety check of the ls action output.
+                            local todo=$( \
+                                TODOTXT_VERBOSE=0 todo.sh -@ -+ -p -x command ls "^ *${cur} " | \
+                                sed -e 's/^ *[0-9]\+ //' -e 's/\((.) \)[0-9]\{2,4\}-[0-9]\{2\}-[0-9]\{2\} /\1/' \
+                                    -e 's/\([xX] \)\([0-9]\{2,4\}-[0-9]\{2\}-[0-9]\{2\} \)\{1,2\}/\1/' \
+                                    -e 's/[[:space:]]*$//' \
+                                    -e '1q' \
+                            )
                             # Append task text as a shell comment. This
                             # completion can be a safety check before a
                             # destructive todo.txt operation.
-                            [ "$item" ] && COMPREPLY[0]="$cur # $item"
+                            [ "$todo" ] && COMPREPLY[0]="$cur # $todo"
                             return 0
                         else
                             return 0