diff options
Diffstat (limited to 'Functions/Calendar/calendar_scandate')
| -rw-r--r-- | Functions/Calendar/calendar_scandate | 124 |
1 files changed, 86 insertions, 38 deletions
diff --git a/Functions/Calendar/calendar_scandate b/Functions/Calendar/calendar_scandate index 4ae2ae606..b3a583705 100644 --- a/Functions/Calendar/calendar_scandate +++ b/Functions/Calendar/calendar_scandate @@ -23,6 +23,19 @@ # from 1900 to 2099 inclusive are matched. # - Although timezones are parsed (complicated formats may not be recognized), # they are then ignored; no time adjustment is made. +# - Embedding of times within dates (e.g. "Wed Jun 16 09:30:00 BST 2010") +# causes horrific problems because of the combination of the many +# possible date and time formats to match. The approach taken +# here is to match the time, remove it, and see if the nearby text +# looks like a date. The problem is that the time matched may not +# be that associated with the date, in which case the time will be +# ignored. To minimise this, when the argument "-a" is given to +# anchor the date/time to the start of the line, we never look +# beyond a newline. So if any date/time strings in the text +# are on separate lines the problem is avoided. +# - If you feel sophisticated enough and wish to avoid any ambiguity, +# you can use RFC 2445 date/time strings, for example 20100601T170000. +# These are parsed in one go. # # The following give some obvious examples; users finding here # a format they like and not subject to vagaries of style may skip @@ -136,7 +149,7 @@ # In this case absolute dates are ignored. emulate -L zsh -setopt extendedglob +setopt extendedglob # xtrace zmodload -i zsh/datetime || return 1 @@ -145,7 +158,7 @@ zmodload -i zsh/datetime || return 1 # relatively logical dates like 2006/09/19:14:27 # don't allow / before time ! the above # is not 19 hours 14 mins and 27 seconds after anything. -local tschars="[-,:[:space:]]" +local tschars="[-,:[:blank:]]" # start pattern for time when anchored local tspat_anchor="(${tschars}#)" # ... when not anchored @@ -175,9 +188,10 @@ local repat="(|s)(|${schars}*)" # We may need some completely different heuristic. local monthpat="(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)[a-z]#" integer daysecs=$(( 24 * 60 * 60 )) +local d="[[:digit:]]" integer year year2 month month2 day day2 hour minute second then nth wday wday2 -local opt line orig_line mname MATCH MBEGIN MEND tz test +local opt line orig_line mname MATCH MBEGIN MEND tz test rest_line local -a match mbegin mend # Flags that we found a date or a time (maybe a relative time) integer date_found time_found @@ -237,7 +251,7 @@ while getopts "aAdmrR:st" opt; do done shift $(( OPTIND - 1 )) -line=$1 orig_line=$1 +line=$1 local dspat dspat_noday tspat if (( anchor )); then @@ -250,11 +264,20 @@ if (( anchor )); then # We'll test later if the time is associated with the date. tspat=$tspat_noanchor fi + # We can save a huge amount of grief (I've discovered) if when + # we're anchored to the start we ignore anything after a newline. + # However, don't do this if we're anchored to the end. The + # match should fail if there are extra lines in that case. + if [[ anchor_end -eq 0 && $line = (#b)([^$'\n']##)($'\n'*) ]]; then + line=$match[1] + rest_line=$match[2] + fi else dspat=$dspat_noanchor dspat_noday=$dspat_noanchor tspat=$tspat_noanchor fi +orig_line=$line # Look for a time separately; we need colons for this. # We want to look for the first time to ensure it's associated @@ -268,6 +291,7 @@ fi # To use a case statement we'd need to be able to request non-greedy # matching for a pattern. local rest +# HH:MM:SECONDS am/pm with optional decimal seconds rest=${line#(#ibm)${~tspat}(<0-12>):(<0-59>)[.:]((<0-59>)(.<->|))[[:space:]]#([ap])(|.)[[:space:]]#m(.|[[:space:]]|(#e))} if [[ $rest != $line ]]; then hour=$match[2] @@ -275,7 +299,8 @@ if [[ $rest != $line ]]; then second=$match[5] [[ $match[7] = (#i)p ]] && (( hour <= 12 )) && (( hour += 12 )) time_found=1 -else +fi +if (( time_found == 0 )); then # no seconds, am/pm rest=${line#(#ibm)${~tspat}(<0-12>):(<0-59>)[[:space:]]#([ap])(|.)[[:space:]]#m(.|[[:space:]]|(#e))} if [[ $rest != $line ]]; then @@ -283,37 +308,60 @@ else minute=$match[3] [[ $match[4] = (#i)p ]] && (( hour <= 12 )) && (( hour += 12 )) time_found=1 - else - # no colon, even, but a.m./p.m. indicator - rest=${line#(#ibm)${~tspat}(<0-12>)[[:space:]]#([ap])(|.)[[:space:]]#m(.|[[:space:]]|(#e))} - if [[ $rest != $line ]]; then - hour=$match[2] - minute=0 - [[ $match[3] = (#i)p ]] && (( hour <= 12 )) && (( hour += 12 )) - time_found=1 - else - # 24 hour clock, with seconds - rest=${line#(#ibm)${~tspat}(<0-24>):(<0-59>)[.:]((<0-59>)(.<->|))(.|[[:space:]]|(#e))} - if [[ $rest != $line ]]; then - hour=$match[2] - minute=$match[3] - second=$match[5] - time_found=1 - else - rest=${line#(#ibm)${~tspat}(<0-24>):(<0-59>)(.|[[:space:]]|(#e))} - if [[ $rest != $line ]]; then - hour=$match[2] - minute=$match[3] - time_found=1 - fi - fi - fi + fi +fi +if (( time_found == 0 )); then + # no colon, even, but a.m./p.m. indicator + rest=${line#(#ibm)${~tspat}(<0-12>)[[:space:]]#([ap])(|.)[[:space:]]#m(.|[[:space:]]|(#e))} + if [[ $rest != $line ]]; then + hour=$match[2] + minute=0 + [[ $match[3] = (#i)p ]] && (( hour <= 12 )) && (( hour += 12 )) + time_found=1 + fi +fi +if (( time_found == 0 )); then + # 24 hour clock, with seconds + rest=${line#(#ibm)${~tspat}(<0-24>):(<0-59>)[.:]((<0-59>)(.<->|))(.|[[:space:]]|(#e))} + if [[ $rest != $line ]]; then + hour=$match[2] + minute=$match[3] + second=$match[5] + time_found=1 + fi +fi +if (( time_found == 0 )); then + rest=${line#(#ibm)${~tspat}(<0-24>):(<0-59>)(.|[[:space:]]|(#e))} + if [[ $rest != $line ]]; then + hour=$match[2] + minute=$match[3] + time_found=1 + fi +fi +if (( time_found == 0 )); then + # Combined date and time formats: here we can use an anchor because + # we know the complete format. + (( anchor )) && tspat=$tspat_anchor + # RFC 2445 + rest=${line#(#ibm)${~tspat}(|\"[^\"]##\":)($~d$~d$~d$~d)($~d$~d)($~d$~d)T($~d$~d)($~d$~d)($~d$~d)([[:space:]]#|(#e))} + if [[ $rest != $line ]]; then + year=$match[3] + month=$match[4] + day=$match[5] + hour=$match[6] + minute=$match[7] + second=$match[8] + # signal don't need to take account of time in date... + time_found=2 + date_found=1 + date_start=$mbegin[3] + date_end=$mend[-1] fi fi (( hour == 24 )) && hour=0 -if (( time_found )); then - # time was found +if (( time_found && ! date_found )); then + # time was found; if data also found already, process below. time_start=$mbegin[2] time_end=$mend[-1] # Remove the timespec because it may be in the middle of @@ -331,7 +379,7 @@ if (( time_found )); then (( debug )) && print "line after time: $line" fi -if (( relative == 0 )); then +if (( relative == 0 && date_found == 0 )); then # Date. case $line in # Look for YEAR[-/.]MONTH[-/.]DAY @@ -468,7 +516,7 @@ if (( date_found || (time_ok && time_found) )); then fi line=${line[1,$date_start-1]}${line[$date_end+1,-1]} fi - if (( time_found )); then + if (( time_found == 1 )); then if (( date_found )); then # If we found a time, it must be associated with the date, # or we can't use it. Since we removed the time from the @@ -540,7 +588,7 @@ if (( date_found || (time_ok && time_found) )); then "'$orig_line[time_start,time_end]'" (( date_ok )) && print "Date string: $date_start,$date_end:" \ "'$orig_line[date_start,date_end]'" - print "Remaining line: '$line'" + print "Remaining line: '$line$rest_line'" fi fi fi @@ -722,11 +770,11 @@ if (( relative )); then (( reladd += (hour * 60 + minute) * 60 + second )) typeset -g REPLY (( REPLY = relative_start + reladd )) - [[ -n $setvar ]] && typeset -g REPLY2="$line" + [[ -n $setvar ]] && typeset -g REPLY2="$line$rest_line" return 0 fi return 1 -elif (( ! date_found )); then +elif (( date_found == 0 )); then return 1 fi @@ -748,6 +796,6 @@ fi strftime -s REPLY -r $fmt $nums -[[ -n $setvar ]] && typeset -g REPLY2="$line" +[[ -n $setvar ]] && typeset -g REPLY2="$line$rest_line" return 0 |