最近在整理论文文献，发现Bib格式总是杂乱无章，所以自己写了一个简单实用的Bash脚本，可以快速清理、规范化 .bib 文献文件。
✨ 特点：
自动去除重复文献
统一文献格式，减少格式出错
操作简单，支持Windows（WSL）和Ubuntu
📌 使用方法：
Windows 用户：通过 WSL 或 Git Bash 运行
Ubuntu 用户：直接在终端运行
直接上代码：
#!/bin/bash
error=1
if [ ! -n "$1" ] || ([ -n "$1" ] && [ $1 == "1" ]); then
error=0
echo "主目录为"${HOME}
echo "注意：有些会议（期刊）的缩写需要手动添加"
read -p "输入bib文件的路径" path
# path="${HOME}/123.bib"
if [ ! -f "$path" ]; then
echo "指定路径$path不存在！"
exit
fi
line_nb=$(sed -n '$=' "$path")
left_path="${path%/*}"
right_path="${path##*/}"
new_path=$left_path/standardized_$right_path
temp_path=$left_path/temp_$right_path
if [ -f "$new_path" ]; then
rm -rf $new_path
fi
touch $new_path
function include_line() {
# 判断字符串是否是包含完整的行
res=0
if [[ "$read_line" =~ "@" ]]; then
res1=$(echo "$read_line" | grep "@")
res2=$(echo "$read_line" | grep "{")
res3=$(echo "$read_line" | grep ",")
if [ "$res1" != "" ] && [ "$res2" != "" ] && [ "$res3" != "" ]; then
res=1
fi
elif [ "$read_line" == "}," ]; then
res=3
else
res4=$(echo "$read_line" | grep "= \+{")
res4_=$(echo "$read_line" | grep "={")
res5=$(echo "$read_line" | grep "} \+,")
res5_=$(echo "$read_line" | grep "},")
if [ "$res4$res4_" != "" ] && [ "$res5$res5_" != "" ]; then
res=2
fi
fi
}
function pre_process() {
echo "正在对bib文件进行预处理"
if [ -f "$temp_path" ]; then
rm -rf $temp_path
fi
touch $temp_path
read_line=""
for ((i = 1; i <= $line_nb; i++)); do
temp_line=$(sed -n ${i}p "$path")
# echo "temp_line=$temp_line"
read_line="$read_line$temp_line"
read_line=$(echo "$read_line" | sed 's/\t//g')
temp_line=$(echo "$read_line" | sed 's/ //g')
if [ "$temp_line" != "" ]; then
read_line=$(echo "$read_line" | awk '{gsub(/^\+|\+$/,"");print}')
read_line=$(echo "$read_line" | sed 's/}$/},/g')
# echo "read_line=$read_line"
while :; do
include_line $read_line
# echo $res
# 将完整部分写入临时文件
if [ $res -eq 1 ]; then
write_line="${read_line%%,*}"
read_line="${read_line#*,}"
echo "$write_line," >>$temp_path
elif [ $res -eq 2 ]; then
read_line=$(echo "$read_line" | sed 's/},/@/g')
read_line=$(echo "$read_line" | sed 's/} \+,/@/g')
write_line="${read_line%%@*}"
read_line="${read_line#*@}"
read_line=$(echo "$read_line" | sed 's/@/},/g')
echo "$write_line}," >>$temp_path
elif [ $res -eq 3 ]; then
echo "$read_line" >>$temp_path
read_line=""
else
break
fi
done
fi
done
path=$temp_path
}
function check_symbol() {
# 检查常规符号
curr_line=$(echo "$curr_line" | sed 's/- \+/-/g')
curr_line=$(echo "$curr_line" | sed 's/ \+-/-/g')
curr_line=$(echo "$curr_line" | sed 's/{ \+/{/g')
curr_line=$(echo "$curr_line" | sed 's/ \+}/}/g')
curr_line=$(echo "$curr_line" | sed 's/(/ (/g')
curr_line=$(echo "$curr_line" | sed 's/( \+/(/g')
curr_line=$(echo "$curr_line" | sed 's/ \+)/)/g')
curr_line=$(echo "$curr_line" | sed 's/ \+,/,/g')
curr_line=$(echo "$curr_line" | sed 's/,/, /g')
curr_line=$(echo "$curr_line" | sed 's/ \+$//g')
curr_line=$(echo "$curr_line" | sed 's/ \+:/:/g')
# 多空格归一化
curr_line=$(echo "$curr_line" | sed 's/ \+/ /g')
}
function adjust_title_journal() {
# 调整title或booktitle或journal的格式
curr_line=$(echo "$curr_line" | sed 's/}$/},/g')
curr_line=$(echo "$curr_line" | sed 's/{//g')
curr_line=$(echo "$curr_line" | sed 's/}//g')
curr_line=$(echo "$curr_line" | sed 's/\\//g')
curr_line=$(echo "$curr_line" | sed 's/\$//g')
curr_line=$(echo "$curr_line" | sed 's/=/={{/g')
curr_line=$(echo "$curr_line" | sed 's/,$/}},/g')
# 处理title中的^
curr_line=$(echo "$curr_line" | sed 's/\^ \+/\\\^{}/g')
# 处理title中的&
curr_line=$(echo "$curr_line" | sed 's/&/\\\&/g')
check_symbol
}
function special_case() {
# 介（连）词小写和专有名词大写
curr_line=$(echo "$curr_line" | sed 's/: \+/:/g')
curr_line=$(echo "$curr_line" | sed 's/ Of / of /g')
curr_line=$(echo "$curr_line" | sed 's/ For / for /g')
curr_line=$(echo "$curr_line" | sed 's/ From / from /g')
curr_line=$(echo "$curr_line" | sed 's/ Between / between /g')
curr_line=$(echo "$curr_line" | sed 's/ With / with /g')
curr_line=$(echo "$curr_line" | sed 's/ And / and /g')
curr_line=$(echo "$curr_line" | sed 's/ The / the /g')
curr_line=$(echo "$curr_line" | sed 's/ At / at /g')
curr_line=$(echo "$curr_line" | sed 's/-A-/-a-/g')
curr_line=$(echo "$curr_line" | sed 's/-An-/-an-/g')
curr_line=$(echo "$curr_line" | sed 's/ A / a /g')
curr_line=$(echo "$curr_line" | sed 's/ An / an /g')
curr_line=$(echo "$curr_line" | sed 's/ By / by /g')
curr_line=$(echo "$curr_line" | sed 's/ In / in /g')
curr_line=$(echo "$curr_line" | sed 's/ To / to /g')
curr_line=$(echo "$curr_line" | sed 's/ Into / into /g')
curr_line=$(echo "$curr_line" | sed 's/ On / on /g')
curr_line=$(echo "$curr_line" | sed 's/ Over / over /g')
curr_line=$(echo "$curr_line" | sed 's/ Empowered / empowered /g')
curr_line=$(echo "$curr_line" | sed 's/ Of / of /g')
curr_line=$(echo "$curr_line" | sed 's/:/: /g')
curr_line=$(echo "$curr_line" | sed 's/Et Al./et al./g')
curr_line=$(echo "$curr_line" | sed "s/'S/'s/g")
curr_line=$(echo "$curr_line" | sed 's/Gpu/GPU/g')
curr_line=$(echo "$curr_line" | sed 's/Cpu/CPU/g')
curr_line=$(echo "$curr_line" | sed 's/Dnn/DNN/g')
curr_line=$(echo "$curr_line" | sed 's/Arm/ARM/g')
}
function publisher() {
# 设置会议（期刊）的出版社
curr_line=$(echo "$curr_line" | sed 's/Ieee/IEEE/g')
curr_line=$(echo "$curr_line" | sed 's/Acm/ACM/g')
curr_line=$(echo "$curr_line" | sed 's/Usenix/USENIX/g')
curr_line=$(echo "$curr_line" | sed 's/(IEEE/(/g')
curr_line=$(echo "$curr_line" | sed 's/(ACM/(/g')
curr_line=$(echo "$curr_line" | sed 's/(USENIX/(/g')
curr_line=$(echo "$curr_line" | sed 's/( \+/(/g')
}
function check_type() {
# 检查引用的类型
echo "从"$start_line"行到"$end_line"行"
for ((j = $start_line; j <= $end_line; j++)); do
line=$(sed -n ${j}p "$path")
line=$(echo "$line" | sed 's/\t//g')
temp=$(echo "$line" | sed 's/ //g')
if [ "$temp" != "" ]; then
# 根据@判断类型
type="misc"
if [ $j -eq $start_line ]; then
if [[ $line =~ "@misc" ]] || [[ $line =~ "@Misc" ]] || [[ $line =~ "@MISC" ]]; then
type="misc"
break
elif [[ $line =~ "@book" ]] || [[ $line =~ "@Book" ]] || [[ $line =~ "@BOOK" ]] ||
[[ $line =~ "@inbook" ]] || [[ $line =~ "@Inbook" ]] || [[ $line =~ "@INBOOK" ]] ||
[[ $line =~ "@incollection" ]] || [[ $line =~ "@Incollection" ]] || [[ $line =~ "@INCOLLECTION" ]]; then
type="book"
break
elif [[ $line =~ "@proceedings" ]] || [[ $line =~ "@Proceedings" ]] || [[ $line =~ "@PROCEEDINGS" ]] ||
[[ $line =~ "@inproceedings" ]] || [[ $line =~ "@Inproceedings" ]] || [[ $line =~ "@INPROCEEDINGS" ]]; then
type="conference"
break
else
type="journal"
fi
else
temp=$(echo $line | sed 's/ //g')
if [ "$temp" != "}" ]; then
# 根据字段判断类型
equal_left="${line%=*}"
if [[ $equal_left =~ "booktitle" ]] || [[ $equal_left =~ "journal" ]]; then
equal_right="${line#*=}"
if [[ $equal_right =~ "symposium" ]] || [[ $equal_right =~ "Symposium" ]] ||
[[ $equal_right =~ "conference" ]] || [[ $equal_right =~ "Conference" ]] ||
[[ $equal_right =~ "workshop" ]] || [[ $equal_right =~ "Workshop" ]]; then
type="conference"
break
elif [[ $equal_right =~ "journal" ]] || [[ $equal_right =~ "Journal" ]] ||
[[ $equal_right =~ "transactions" ]] || [[ $equal_right =~ "Transactions" ]]; then
type="journal"
break
elif [[ $equal_right =~ "proceedings" ]] || [[ $equal_right =~ "Proceedings" ]]; then
type="conference"
break
fi
if [[ $equal_left =~ "booktitle" ]]; then
type="conference"
break
elif [[ $equal_left =~ "journal" ]]; then
type="journal"
break
fi
fi
fi
fi
fi
done
echo $type
}
function remove_extra_blank() {
# 去除字符串中=和{}控制下的多余空格
if [ $k -eq $start_line ]; then
left_line="${curr_line%%\{*}"
right_line="${curr_line#*\{}"
left_line=$(echo $left_line | sed 's/ //g')
right_line=$(echo $right_line | sed 's/ //g')
curr_line=$left_line{$right_line
else
temp=$(echo $curr_line | sed 's/ //g')
if [ "$temp" != "}" ]; then
left_line="${curr_line%%\{*}"
temp_line="${curr_line#*\{}"
mid_line="${temp_line%\}*}"
right_line="${temp_line##*\}}"
left_line=$(echo $left_line | sed 's/ //g')
right_line=$(echo $right_line | sed 's/ //g')
# echo "左"$left_line"中"$mid_line"右"$right_line
curr_line=$left_line{$mid_line}$right_line
fi
fi
}
function judge_more_A_than_a() {
local strings="$1"
local A=0
local a=0
for ((i = 0; i < $(echo ${#strings}); i++)); do
e=$(echo ${strings:$i:1})
if [[ $e = [[:lower:]] ]]; then
a=$(($a + 1))
elif [[ $e = [[:upper:]] ]]; then
A=$(($A + 1))
fi
done
if [ $A -gt $a ]; then
echo "1"
else
echo "0"
fi
}
function rebuild_cite() {
# 重新生成引用
start_line=$1
end_line=$2
check_type
for ((k = $start_line; k <= $end_line; k++)); do
line=$(sed -n ${k}p "$path")
# 去除所有TAB
line=$(echo "$line" | sed 's/\t//g')
# 考虑去除所有空格后的字符串
temp=$(echo "$line" | sed 's/ //g')
if [ "$temp" != "" ]; then
# 去除字符串左右两边的空格
curr_line=$(echo "$line" | awk '{gsub(/^\+|\+$/,"");print}')
remove_extra_blank $start_line $curr_line $k
# 确定@对象的形式
if [[ $curr_line =~ "@" ]]; then
curr_line="${curr_line#*\{}"
if [ $type == "conference" ]; then
echo "@inproceedings{$curr_line" >>$new_path
elif [ $type == "journal" ]; then
echo "@article{$curr_line" >>$new_path
elif [ $type == "book" ]; then
echo "@book{$curr_line" >>$new_path
elif [ $type == "misc" ]; then
echo "@misc{$curr_line" >>$new_path
fi
# 确定title的形式
elif [[ $curr_line =~ "title" ]] && ! [[ $curr_line =~ "booktitle" ]]; then
adjust_title_journal
curr_line=$(echo "$curr_line" | sed 's/\b[a-z]/\U&/g')
curr_line=$(echo "$curr_line" | sed 's/Title={{/title={{/g')
special_case
echo " $curr_line" >>$new_path
# 确定author的形式
elif [[ $curr_line =~ "author" ]]; then
curr_line=$(echo "$curr_line" | sed 's/{//g')
curr_line=$(echo "$curr_line" | sed 's/}//g')
curr_line=$(echo "$curr_line" | sed 's/=/={/g')
curr_line=$(echo "$curr_line" | sed 's/,$/},/g')
check_symbol
echo " $curr_line" >>$new_path
# 确定booktitle/journal的形式
elif [[ $curr_line =~ "booktitle" ]] || [[ $curr_line =~ "journal" ]]; then
if [ $type == "conference" ] || [ $type == "journal" ]; then
curr_line=$(echo "$curr_line" | sed 's/in Proceedings of the//g')
curr_line=$(echo "$curr_line" | sed 's/in Proceedings of//g')
curr_line=$(echo "$curr_line" | sed 's/Proceedings of the//g')
curr_line=$(echo "$curr_line" | sed 's/Proceedings of//g')
if [ $type == "conference" ]; then
curr_line=$(echo $curr_line | sed 's/[iI]nterest /Interestt /g')
curr_line=$(echo $curr_line | sed 's/[-a-zA-Z0-9]\+st / /g')
curr_line=$(echo $curr_line | sed 's/Interestt /Interest /g')
curr_line=$(echo $curr_line | sed 's/ [aA]nd / andd /g')
curr_line=$(echo $curr_line | sed 's/[-a-zA-Z0-9]\+nd / /g')
curr_line=$(echo $curr_line | sed 's/ andd / and /g')
curr_line=$(echo $curr_line | sed 's/[-a-zA-Z0-9]\+th / /g')
curr_line=$(echo $curr_line | sed 's/[0-9]\+//g')
fi
adjust_title_journal
curr_line=$(echo "$curr_line" | sed 's/\b[a-z]/\U&/g')
special_case
publisher
if [ $type == "conference" ]; then
# 删除,之后的多余字段并添加,之后的会议缩写
curr_line=$(echo "$curr_line" | sed "s/USA//g")
abbrev0=$(echo "${curr_line#*,}")
abbrev=$abbrev0
while [[ "$abbrev" =~ "," ]]; do
temp_abbrev=$(echo "${abbrev%%,*}")
res=$(judge_more_A_than_a $temp_abbrev)
if [ "$res" == "1" ]; then
break
fi
abbrev=$(echo "${abbrev#*,}")
done
if [[ "$abbrev0" =~ "," ]]; then
if [[ "$abbrev" =~ "," ]]; then
# echo $abbrev
# echo $temp_abbrev
curr_line=$(echo "$curr_line" | sed "s#$abbrev#@#g")
curr_line=$(echo "$curr_line" | sed 's/,@/}},/g')
temp_abbrev=$(echo "$temp_abbrev" | sed "s#^ \+##g")
curr_line=$(echo "$curr_line" | sed "s#}},# ($temp_abbrev)}},#g")
else
curr_line=$(echo "$curr_line" | sed "s#$abbrev0#@#g")
curr_line=$(echo "$curr_line" | sed 's/,@/}},/g')
fi
fi
# 删除-IEEE之前的多余字段并添加-IEEE之前的会议缩写
curr_line=$(echo "$curr_line" | sed 's/-IEEE/@/g')
if [[ $curr_line =~ "@" ]]; then
abbrev=$(echo "${curr_line%@*}")
abbrev=$(echo "$abbrev" | sed 's/IEEE//g')
abbrev=$(echo "$abbrev" | sed 's/[0-9]*//g')
abbrev=$(echo "$abbrev" | sed 's/ //g')
abbrev=$(echo "${abbrev##*\{}")
curr_line=$(echo "$curr_line" | sed 's/={{.*@/={{IEEE/g')
if [ "$abbrev" != "" ]; then
curr_line=$(echo "$curr_line" | sed "s/}},/ ($abbrev)}},/g")
fi
fi
# 调整IEEE和ACM的格式
curr_line=$(echo "$curr_line" | sed 's/IEEE International/IEEE/g')
curr_line=$(echo "$curr_line" | sed 's/IEEE /IEEE International /g')
curr_line=$(echo "$curr_line" | sed 's#IEEE/ACM International#IEEE/ACM#g')
curr_line=$(echo "$curr_line" | sed 's#IEEE/ACM#IEEE/ACM International#g')
fi
# echo "$curr_line"
curr_line="${curr_line#*=}"
if [ $type == "conference" ]; then
curr_line="booktitle=$curr_line"
curr_line=$(echo "$curr_line" | sed 's/={{/={{Proceedings of the /g')
elif [ $type == "journal" ]; then
curr_line="journal=$curr_line"
fi
echo " $curr_line" >>$new_path
else
# echo "删除booktitle/journal"
if [ $k -eq $end_line ]; then
echo "}" >>$new_path
fi
fi
# 确定organization/publisher的形式
elif [[ $curr_line =~ "organization" ]] || [[ $curr_line =~ "publisher" ]]; then
if [ $type == "conference" ] || [ $type == "journal" ]; then
# echo "删除organization/publisher"
if [ $k -eq $end_line ]; then
echo "}" >>$new_path
fi
elif [ $type == "book" ]; then
curr_line=$(echo "$curr_line" | sed 's/{//g')
curr_line=$(echo "$curr_line" | sed 's/}//g')
curr_line=$(echo "$curr_line" | sed 's/=/={/g')
curr_line=$(echo "$curr_line" | sed 's/,*$/},/g')
check_symbol
echo " $curr_line" >>$new_path
fi
elif [[ $curr_line =~ "editor" ]] || [[ $curr_line =~ "series" ]]; then
# echo "删除editor/series"
if [ $k -eq $end_line ]; then
echo "}" >>$new_path
fi
else
check_symbol
# 确定该引用最后一行的形式
if [ $k -eq $end_line ]; then
temp=$(echo $curr_line | sed 's/ //g')
if [[ $temp =~ "}}" ]]; then
temp_line="${curr_line%%\}*}"
echo " $temp_line}" >>$new_path
echo "}" >>$new_path
else
echo "}" >>$new_path
fi
else
echo " $curr_line" >>$new_path
fi
fi
fi
done
echo >>$new_path
echo
}
if [ -n "$1" ] && [ $1 == "1" ]; then
pre_process
fi
cite_nb=0
old_start=0
start=0
end=0
for ((i = 1; i <= $line_nb; i++)); do
line=$(sed -n ${i}p "$path")
str=$(echo $line | sed 's/ //g')
if ! [[ $line =~ "Encoding" ]] && ! [[ $line =~ "@Comment" ]]; then
if [[ $line =~ "@" ]]; then
let cite_nb++
old_start=$start
start=$i
if [ $end -ne 0 ]; then
rebuild_cite $old_start $end
fi
echo "获取第"$cite_nb"个引用"
elif [ "$str" != "" ]; then
end=$i
fi
fi
done
if [ $start -ne 0 ] && [ $end -ne 0 ]; then
rebuild_cite $start $end
fi
sed -i '$d' $new_path
# 检查}后的逗号
line_nb=$(sed -n '$=' "$new_path")
let line_nb=line_nb-1
for ((i = 1; i <= $line_nb; i++)); do
line=$(sed -n ${i}p "$new_path")
let j=i+1
next=$(sed -n ${j}p "$new_path")
if [ "$next" == "}" ]; then
new_line=$(echo "*$line" | sed 's/},/}/g')
sed -i "${i}c${new_line}" $new_path
fi
done
sed -i "s/*//g" $new_path
echo "已经生成"$new_path
rm -rf $temp_path
fi
if [ $error == "1" ]; then
echo "./bibtex.sh 快速模式（适用每个字段单独占据一行的情况）"
echo "./bibtex.sh 1 完全模式（适用每个字段占据多行或多个字段占据一行的情况）"
exit
fi
# TODO 解决会议/期刊的缩写问题
扫一扫
2011
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
