FFmpeg/tools/patcheck

#!/bin/sh

# if no argument provided, write stdin to a file and re-run the script
if [ $# = 0 ]; then
    cat > patcheck.stdout
    $0 patcheck.stdout
    rm -f patcheck.stdout
    exit
fi

GREP=grep
EGREP=egrep
TMP=patcheck.tmp
OPT="-nH"
#FILES=$($GREP '^+++' $* | sed 's/+++ //g')

echo patCHeck 1e10.0
echo This tool is intended to help a human check/review patches. It is very far from
echo being free of false positives and negatives, and its output are just hints of what
echo may or may not be bad. When you use it and it misses something or detects
echo something wrong, fix it and send a patch to the ffmpeg-devel mailing list.
echo License: GPL, Author: Michael Niedermayer

ERE_PRITYP='(unsigned *|)(char|short|long|int|long *int|short *int|void|float|double|(u|)int(8|16|32|64)_t)'
ERE_TYPES='(const|static|av_cold|inline| *)*('$ERE_PRITYP'|[a-zA-Z][a-zA-Z0-9_]*)[* ]{1,}[a-zA-Z][a-zA-Z0-9_]*'
ERE_FUNCS="$ERE_TYPES"' *\('

hiegrep(){
    arg="$1"
    msg="$2"
    shift 2
    $GREP $OPT '^+' $* | $GREP -v ':+++'| $EGREP --color=always -- "$arg"> $TMP && printf "\n$msg\n"
    cat $TMP
}

hiegrep2(){
    arg="$1"
    varg="$2"
    msg="$3"
    shift 3
    $GREP $OPT '^+' $* | $GREP -v ':+++' | $EGREP -v -- "$varg" | $EGREP --color=always -- "$arg" > $TMP && printf "\n$msg\n"
    cat $TMP
}

hiegrep 'static[^(]*\*[a-zA-Z_]*\[' 'pointer array is not const' $*
hiegrep '[[:space:]]$'    'trailing whitespace' $*
hiegrep "$(echo x | tr 'x' '\t')"         'tabs' $*
#hiegrep ':\+$'          'Empty lines' $*
hiegrep ';;'              'double ;' $*
hiegrep2 '\b_[a-zA-Z0-9_]{1,}' '__(asm|attribute)([^a-zA-Z0-9]|$)' 'reserved identifer' $*
hiegrep '//[-/<\* ]*$'    'empty comment' $*
hiegrep '/\*[-<\* ]*\*/'  'empty comment' $*
hiegrep '(static|inline|const) *\1[^_a-zA-Z]'  'duplicate word' $*
hiegrep '=[-+\*\&] ' 'looks like compound assignment' $*
hiegrep2 '/\*\* *[a-zA-Z0-9].*' '\*/' 'Inconsistently formatted doxygen comment' $*
hiegrep '; */\*\*[^<]' 'Misformatted doxygen comment' $*
hiegrep '//!|/\*!' 'inconsistent doxygen syntax' $*

hiegrep2 '(int|unsigned|static|void)[a-zA-Z0-9 _]*(init|end)[a-zA-Z0-9 _]*\(.*[^;]$' '(av_cold|:\+[^a-zA-Z_])' 'These functions may need av_cold, please review the whole patch for similar functions needing av_cold' $*

hiegrep '\+= *1 *;'     'can be simplified to ++' $*
hiegrep '-= *1 *;'      'can be simplified to --' $*
hiegrep '((!|=)= *(0|NULL)[^0-9a-z]|[^0-9a-z](0|NULL) *(!|=)=)' 'x==0 / x!=0 can be simplified to !x / x' $*

$EGREP $OPT '^\+ *(const *|)static' $*| $EGREP --color=always '[^=]= *(0|NULL)[^0-9a-zA-Z]'> $TMP && printf '\nuseless 0 init\n'
cat $TMP
hiegrep '# *ifdef * (HAVE|CONFIG)_' 'ifdefs that should be #if' $*

hiegrep '\b(awnser|cant|dont|wont|doesnt|usefull|successfull|occured|teh|alot|wether|skiped|skiping|heigth|informations|colums|loosy|loosing|ouput|seperate|preceed|upto|paket|posible|unkown|inpossible|dimention|acheive|funtions|overriden|outputing|seperation|initalize|compatibilty|bistream|knwon|unknwon|choosen|additonal|gurantee|availble|wich|begining|milisecond|missmatch|threshhold|accomodate|processsing|substract|auxillary|coefficents|neccesary|precison)\b' 'common typos' $*

hiegrep 'av_log\( *NULL' 'Missing context in av_log' $*
hiegrep '[^sn]printf' 'Please use av_log' $*
hiegrep '\bmalloc' 'Please use av_malloc' $*
hiegrep '\) *av_malloc' 'useless casts' $*
hiegrep ':\+ *'"$ERE_PRITYP"' *inline' 'non static inline or strangely ordered inline+static' $*
hiegrep "$ERE_FUNCS"' *\)' 'missing void' $*
hiegrep '(sprintf|strcat|strcpy)' 'Possible security issue, make sure this is safe or use snprintf/av_strl*' $*
hiegrep '/ *(2|4|8|16|32|64|128|256|512|1024|2048|4096|8192|16384|32768|65536)[^0-9]' 'divide by 2^x could use >> maybe' $*
hiegrep '#(el|)if *(0|1)' 'useless #if' $*
hiegrep 'if *\( *(0|1) *\)' 'useless if()' $*
hiegrep '& *[a-zA-Z0-9_]* *\[ *0 *\]' 'useless & [0]' $*
hiegrep '(\( *[0-9] *(&&|\|\|)|(&&|\|\|) *[0-9] *\))' 'overriding condition' $*
hiegrep '(:\+|,|;)( *|static|\*)*'"$ERE_PRITYP"' *\*( |\*)*(src|source|input|in[^a-z])' 'missing const?' $*
hiegrep '(:\+|,|;)( *|static|\*)*'"$ERE_PRITYP"' *(src|source|input|in)([0-9A-Z_][0-9A-Za-z_]*){1,} *\[' 'missing const (test2)?' $*
hiegrep ' *static *'"$ERE_FUNCS"'[^)]*\);' 'static prototype, maybe you should reorder your functions' $*
hiegrep '@file: *[a-zA-Z0-9_]' 'doxy filetag with filename can in the future cause problems when forgotten during a rename' $*
hiegrep '\bassert' 'Please use av_assert0, av_assert1 or av_assert2' $*

hiegrep2 '\.long_name *=' 'NULL_IF_CONFIG_SMAL' 'missing NULL_IF_CONFIG_SMAL' $*
hiegrep2 '\.pix_fmts *= *\(' 'const' 'missing const for pix_fmts array' $*
hiegrep2 '\.sample_fmts *= *\(' 'const' 'missing const for sample_fmts array' $*
hiegrep2 '\.supported_framerates *= *\(' 'const' 'missing const for supported_framerates array' $*
hiegrep2 '\.channel_layouts *= *\(' 'const' 'missing const for channel_layouts array' $*

#$EGREP $OPT '^\+.*const ' $*| $GREP -v 'static'> $TMP && printf '\nnon static const\n'
#cat $TMP

hiegrep2 "$ERE_TYPES" '(static|av_|ff_|typedef|:\+[^a-zA-Z_])' 'Non static with no ff_/av_ prefix' $*

hiegrep ':\+[^}#]*else' 'missing } prior to else' $*
hiegrep '(if|while|for)\(' 'missing whitespace between keyword and ( (feel free to ignore)' $*
hiegrep '(else|do){'       'missing whitespace between keyword and { (feel free to ignore)' $*
hiegrep '}(else|while)'    'missing whitespace between } and keyword (feel free to ignore)' $*

#FIXME this should print the previous statement maybe
hiegrep ':\+  *{ *$' '{ should be on the same line as the related previous statement' $*


rm $TMP
for i in $($GREP -H '^+.*@param' $*| sed 's/^\([^:]*\):.*@param\(\[.*\]\|\) *\([a-zA-Z0-9_]*\) .*$/\1:\3/') ; do
    doxpar=$(echo $i | sed 's/^.*:\(.*\)$/\1/')
    file=$(echo $i | sed 's/^\([^:]*\):.*$/\1/')
    $GREP " *$doxpar *[),]" $file | $GREP -v '@param' >/dev/null || $GREP --color=always "@param *$doxpar" $file >>$TMP
done
if test -e $TMP ; then
    printf '\nmismatching doxy params\n'
    cat $TMP
fi

$EGREP -B2 $OPT '^(\+|) *('"$ERE_TYPES"'|# *define)' $* | $EGREP -A2 --color=always '(:|-)\+[^/]*/(\*([^*]|$)|/([^/]|$))' > $TMP && printf "\n Non doxy comments\n"
cat $TMP

rm $TMP
for i in \
    $($EGREP -H '^\+ *'"$ERE_TYPES" $*  |\
    $GREP -v '(' | $EGREP -v '\Wgoto\W' |\
    xargs -d '\n' -n 1 |\
    $GREP -o '[* ][* ]*[a-zA-Z][0-9a-zA-Z_]* *[,;=]' |\
    sed 's/.[* ]*\([a-zA-Z][0-9a-zA-Z_]*\) *[,;=]/\1/') \
    ; do
    echo $i | $GREP '^NULL$' && continue
    $EGREP $i' *(\+|-|\*|/|\||&|%|)=[^=]' $* >/dev/null || echo "possibly never written:"$i >> $TMP
    $EGREP '(=|\(|return).*'$i'(==|[^=])*$'    $* >/dev/null || echo "possibly never read   :"$i >> $TMP
    $EGREP -o $i' *((\+|-|\*|/|\||&|%|)=[^=]|\+\+|--) *(0x|)[0-9]*(;|)'   $* |\
           $EGREP -v $i' *= *(0x|)[0-9]{1,};'>/dev/null || echo "possibly constant     :"$i >> $TMP
done
if test -e $TMP ; then
    printf '\npossibly unused variables\n'
    cat $TMP
fi

$GREP '^+++ .*Changelog' $* >/dev/null || printf "\nMissing changelog entry (ignore if minor change)\n"

cat $* | $GREP -v '^-' | tr '\n' '@' | $EGREP --color=always -o '(fprintf|av_log|printf)\([^)]*\)[+ ;@]*\1'  >$TMP && printf "\nMergeable calls\n"
cat $TMP | tr '@' '\n'

cat $* | tr '\n' '@' | $EGREP --color=always -o '\+ *if *\( *([A-Za-z0-9_]*) *[<>]=? *[0-9]* *\) * \1 *= *[0-9]* *;[ @\\+]*else *if *\( *\1 *[<>]=? *[0-9]* *\) *\1 *= *[0-9]* *;'  >$TMP && printf "\nav_clip / av_clip_uint8 / av_clip_int16 / ...\n"
cat $TMP | tr '@' '\n'

cat $* | tr '\n' '@' | $EGREP --color=always -o '\+ *if *\( *([A-Za-z0-9_]*) *[<>]=? *([A-Za-z0-9_]*) *\)[ @\\+]*(\1|\2) *= *(\1|\2) *;'  >$TMP && printf "\nFFMIN/FFMAX\n"
cat $TMP | tr '@' '\n'

cat $* | tr '\n' '@' | $EGREP --color=always -o '\+ *if *\( *([A-Za-z0-9_]*) *\)[ @\\+]*av_free(p|) *\( *(&|) *\1[^-.]'  >$TMP && printf "\nav_free(NULL) is safe\n"
cat $TMP | tr '@' '\n'

cat $* | tr '\n' '@' | $EGREP --color=always -o '[^a-zA-Z0-9_]([a-zA-Z0-9_]*) *= *av_malloc *\([^)]*\)[ @;\\+]*memset *\( *\1'  >$TMP && printf "\nav_mallocz()\n"
cat $TMP | tr '@' '\n'


# does not work
#cat $* | tr '\n' '@' | $EGREP -o '[^a-zA-Z_0-9]([a-zA-Z][a-zA-Z_0-9]*) *=[^=].*\1' | $EGREP -o '[^a-zA-Z_0-9]([a-zA-Z][a-zA-Z_0-9]*) *=[^=].*\1 *=[^=]'  >$TMP && printf "\nPossibly written 2x before read\n"
#cat $TMP | tr '@' '\n'

exit

TODO/idea list:

for all demuxers & muxers
    $EGREP for "avctx->priv_data"

vertical align =
/* and * align
arrays fitting in smaller types
variables written to twice with no interspaced read
memset(block, 0, 6*64*sizeof(int16_t)); -> clear_blocks
check existence of long_name in AVCodec
check that the patch does not touch codec & (de)muxer layer at the same time ->split

write a regression test containing at least a line that triggers each warning once