（转载）基于Bash命令行的百度云上传下载工具

原文链接：http://hi.baidu.com/meoow/item/aef5814bbd5be3e1bcf451e9

这是我根据百度云PCS的API写的一个基于bash的命令行工具，

使用了curl, awk, sed, xxd, bash，没有使用任何其他的语言解释器做辅助

目前，上传下载，文件列表，文件信息，创建文件夹，删除文件都没问题

复制和移动总是返回参数错误，实在找不出来哪里跟API不一致了，暂时放弃。

另外离线下载相关的API也都添加了，但是因为现在百度云不能申请，我是用的是bypy的client_id，

不知道它的离线下载相关权限没有抑或其他问题，添加任务返回500。

不过总的来说基本上对于一般使用百毒云是够用了

#!/bin/bash

# Baidu Yun Command Line Interface

# Depends: bash, curl, grep, awk, sed, xxd

# (Thay are basicly builtin tools of any *nix system.)

# Additionally, fastupload depends: head, wc, md5sum or md5, cksum

# (Which are also builtin tools)

# issues:

# move, copy and all offline-download related commnads do not work

# the latters are probably prohibited by baidu

# or the current bypy client has no permission for them.

#### Variables ####

# This three to be changed to your own code #

TOKEN='00.00000000000000000000000000000000.0000000.0000000000.000000000000000000'

# baidu has prohibited registering new client id

# the client_id is borrowed from bypy (Another great baidu yun cli tool written in python)

CLIENT_ID=q8WE4EpCsau1oS0MplgMKNBn

# you should at least login pan.baidu.com from browser once,

# then you can obtain the BDUSS code in the Cookie.

BDUSS='AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA;'

########################################

pathprefix='/apps/bypy/'

ACCEPT='text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8'

ACCPET_LANGUAGE='zh-cn,zh;q=0.8,en-us;q=0.5,en;q=0.3'

ACCPET_ENCODING='identity'

CACHE_CONTROL='no-cache'

CONNECTION='keep-alive'

DNT=

HOST='pcs.baidu.com'

ORIGIN='http://pan.baidu.com'

PRAGMA='no-cache'

REFERER='http://pan.baidu.com/disk/home'

UA='Mozilla/5.0 (X11; Linux; rv:5.0) Gecko/5.0 Firefox/5.0'

CURL_DEFAULT_ARGS=(

    -H "Accept: $ACCEPT"

    -H "Accept-Language: $ACCPET_LANGUAGE"

    -H "Cache-Control: $CACHE_CONTROL"

    -H "Connection: $CONNECTION"

    -H "DNT: $DNT"

    -H "Host: $HOST"

    -H "Origin: $ORIGIN"

    -H "Pragma: $PRAGMA"

    -H "Referer: $REFERER"

    -H "Accept-Encoding: $ACCPET_ENCODING"

    -A "$UA"

)

####################

#### Defining functions ####

# encoding character to %xx format

urlencode() {

    local length="${#1}"

    local c

    for (( i = ; i < length; i++ )); do

    c="${1:i:1}"

    case $c in

    testa-zA-Z0-.~_-) printf "%s" "$c" ;;

    *) printf "%s" "$c" | xxd -p -c1 | while read x;do printf "%%%s" "$x";done

    esac

done

}

# for *basic* regexp syntax surrounded by double quotes

regexescape() {

local len="${#1}"

local c

for(( i=;i<len;i++ ));do

c="${1:i:1}"

case "$c" in

test*\\/^$|']'|'['|'-')

    printf "\\\\%s" "$c"

;;

*)

    printf "%s" "$c"

    esac

done

}

# baidu yun authorize

auth() {

local code

echo "http://openapi.baidu.com/oauth/2.0/authorize?response_type=code&client_id=$CLIENT_ID&redirect_uri=oob&scope=basic%20netdisk&"

code=$(curl -Ls "http://openapi.baidu.com/oauth/2.0/authorize?response_type=code&client_id=$CLIENT_ID&redirect_uri=oob&scope=basic%20netdisk&" -H "Cookie: BDUSS=$BDUSS"|grep -o 'readonly value="[^"]*"'|grep -o '[[:xdigit:]]\{32\}')

curl -Ls "https://bypy-tianze.rhcloud.com/auth?code=$code&redirect_uri=oob&"|

grep -o '"[^"]\{1,\}":"[^"]\{1,\}"'|sed -n '/access_token/s/^.*:"\([[:xdigit:].-]\{1,\}\)"$/\1/p'

}

# print help

usage() {

cat <<_EOF_

    $ <command > [arguments...]

    Command:

    auth

    upload    <FILE > [PATH] [overwrite|newcopy]

    fastupload    <FILE > [PATH] [overwrite|newcopy]

    download  <PATH > [SAVEDIR] [SAVENAME]

    list      [PATH]

    mkdir     <PATH >

    meta      <PATH >

    move      <FROM > <TO>

    copy      <FROM > <TO>

    delete    <PATH >

    search    <PATH > <WORD>

    add       <URI > [PATH]

    query     <TASKID >

    tasks

    cancel    <TASKID >

    trash

    retore    <FILEID >

    _EOF_

}

# JSON pretty printer

tokenize () {

    local GREP

    local ESCAPE

    local CHAR

    GREP='egrep -ao'

    ESCAPE='(\\[^u[:cntrl:]]|\\u[0-9a-fA-F]{4})'

    CHAR='[^[:cntrl:]"\\]'

    local STRING="\"$CHAR*($ESCAPE$CHAR*)*\""

    local NUMBER='-?(0|[1-9][0-9]*)([.][0-9]*)?([eE][+-]?[0-9]*)?'

    local KEYWORD='null|false|true'

    local SPACE='[[:space:]]+'

    $GREP "$STRING|$NUMBER|$KEYWORD|$SPACE|." | egrep -v "^$SPACE$"

}

printw() {

    local n="$1"

    if test[ "$n" =~ [-][-]* ];then

        :

    else

        return

    fi

    for((i=;i<n;i++));do

    printf " "

done

}

pprint() {

local indent=

local value=

while read x; do

    case "$x" in

    '{'|'[')

        if ((value));then

            printf " "

            value=

        else

            printw $indent

        fi

        ((indent++))

        printf "%s\n" "$x"

    ;;

    '}'|']')

        value=

        ((indent >  && indent--))

        echo

        printw $indent

        printf "%s" "$x"

    ;;

    ',')

        value=

        echo

    ;;

    ':')

        value=

        printf "%s" "$x"

    ;;

    *)

        if ((value));then

            value=

            printf " "

        else

            printw $indent

        fi

        printf "%s" "$x"

    ;;

    esac

done

}

# compact json output (parse output from pprint only)

compact() {

# arguments: -f -t -b -m -s

# -f no fs_id

# -t no ctime and mtime

# -b no brackets

# -m no md5

# -s no size

#default is to remove double quotes

local sedcmd='s/"\([^"]*\)"/\1/g;'

while getopts :ftbms arg;do

    case "$arg" in

    f)

        sedcmd="$sedcmd"'/^ *fs_id/d;'

    ;;

    t)

        sedcmd="$sedcmd"'/^ *ctime/d;/^ *mtime/d;'

    ;;

    m)

        sedcmd="$sedcmd"'/^ *md5:/d;'

    ;;

    s)

        sedcmd="$sedcmd"'/^ *size:/d;'

    ;;

    b) sedcmd="$sedcmd"'/^ *{/d;s/^ *}$//;/^ *\[$/d;/^ *\]/d;s/: \[$/:/;'

    ;;

    esac

done

sed "$sedcmd"

}

autounit() {

awk -v G=$((**)) -v M=$((**)) -v K= '

{

    if($~/^ *"?size"?: [[:digit:]]{,}/){

        if($> =G){

            size=sprintf("%0.1fG",$/G)

        } else if($> =M){

            size=sprintf("%0.1fM",$/M)

        } else if($> =K){

            size=sprintf("%0.1fK",$/K)

        } else {

            size=$

        }

        sub($,size)

    }

    print

}

'

}

crc32() {

cksum -o3 < "$1" |

while read x y;do

    echo $x

    break

done

}

if type -t md5sum > /dev/null ;then

MD5='md5sum -b'

else

MD5='md5'

fi

md5() {

MD5 < "$1" |

while read x y;do

    echo "$x"

    break

done

}

md5_256k() {

head -c262144 "$1" | MD5 |

while read x y;do

    echo "$x"

    break

done

}

size() {

wc -c "$1"|

while read x y;do

    echo "$x"

    break

done

}

########################################

#### Parse command line arguments ####

case "$1" in

auth) auth

exit

;;

u|upload)

method=upload

shift

if test[ -z "$1" ];then

    echo "$0 upload <FILE > [PATH] [overwrite|newcopy]" >&

    exit

fi

filepath="$1"

filename="$(basename "$filepath")"

shift

path="${1:-/}/$filename"

shift

ondup="${1:-overwrite}"

;;

f|fastupload)

method=rapidupload

shift

if test[ -z "$1" ];then

    echo "$0 upload <FILE > [PATH] [overwrite|newcopy]" >&

    exit

fi

filepath="$1"

filename="$(basename "$filepath")"

shift

path="${1:-/}/$filename"

shift

ondup="${1:-overwrite}"

;;

d|download)

method=download

shift

if test[ -z "$1" ];then

    echo "$0 download <PATH > [SAVEDIR] [SAVENAME]" >&

    exit

fi

path="$1"

shift

savedir="${1:-.}"

shift

savename="${1:-$(basename "$path")}"

savefile="$savedir/$savename"

;;

mkdir)

req=POST

method=mkdir

shift

if test[ -z "$1" ];then

    echo "$0 mkdir <PATH > " >&

    exit

fi

path="$1"

;;

meta)

req=GET

method=meta

shift

if test[ -z "$1" ];then

    echo "$0 meta <PATH > " >&

    exit

fi

path="$1"

;;

l|list)

method=list

shift

path="${1:-/}"

;;

m|move)

method=move

shift

from="$1"

shift

to="$1"

if test[ -z "$from" || -z "$to" ];then

    echo "$0 move <FROM > <TO> " >&

    exit

fi

;;

c|copy)

method=copy

shift

from="$1"

shift

to="$1"

if test[ -z "$from" || -z "$to" ];then

    echo "$0 copy <FROM > <TO> " >&

    exit

fi

;;

s|search)

method=search

shift

path="$1"

shift

wd="$1"

if test[ -z "$path" || -z "$wd" ];then

    echo "$0 search <PATH > <WORD> " >&

    exit

fi

;;

del|delete)

method=delete

if test[ -z "$1" ];then

    echo "$0 delete <PATH > " >&

    exit

fi

shift

path="$1"

;;

add)

method=add_task

shift

if test[ -z "$1" ];then

    echo "$0 add <LINK > " >&

    exit

fi

link="$1"

shift

path="$1"

;;

query)

method=query_task

shift

if test[ -z "$1" ];then

    echo "$0 query <TASKID > " >&

    exit

fi

taskid="$1"

;;

tasks)

method=list_task

;;

cancel)

method=cancel_task

shift

if test[ -z "$1" ];then

    echo "$0 cancel <TASKID > " >&

    exit

fi

taskid="$1"

;;

trash)

method=listrecycle

;;

restore)

method=restore

shift

if test[ -z "$1" ];then

    echo "$0 restore <FILEID > " >&

    exit

fi

fsid="$1"

;;

h|help)

usage

exit

;;

*)

usage

exit

;;

esac

path="$pathprefix""$path"

from="$pathprefix""$from"

to="$pathprefix""$to"

#### determining request type ####

case $method in

info|download|meta|list|search|generate|diff|streaming|listrecycle)

req=GET

;;

upload|delete|createsuperfile|move|copy|rapidupload|add_task|list_task|query_task|cancel_task|restore|mkdir)

req=POST

;;

esac

##################################

#### determining URI #############

case $method in

upload) uriprefix='https://c.pcs.baidu.com/rest/2.0/pcs/file'

;;

download) uriprefix='https://d.pcs.baidu.com/rest/2.0/pcs/file'

;;

list_task|query_task|cancel_task)

uriprefix='https://pcs.baidu.com/rest/2.0/pcs/services/cloud_dl'

;;

*) uriprefix='https://pcs.baidu.com/rest/2.0/pcs/file'

;;

esac

##################################

#### Building URI depends on specific method ####

uriprefix="$uriprefix?access_token=$TOKEN&method=$method"

case $method in

upload)

uri="$uriprefix&path=$(urlencode "$path")&ondup=$ondup"

;;

download|mkdir|meta|delete)

uri="$uriprefix&path=$(urlencode "$path")"

;;

list)

by=name #time name size

order=asc #asc desc

uri="$uriprefix&path=$(urlencode "$path")&by=$by&order=$order"

;;

move|copy)

uri="$uriprefix"

;;

search)

re=

uri="$uriprefix&path=$(urlencode "$path")&wd=$(urlencode "$wd")&re=$re"

;;

rapidupload)

filelength=$(size "$filepath")

filemd5=$(md5 "$filepath")

fileslicemd5=$(md5_256k "$filepath")

filecrc32=$(crc32 "$filepath")

uri="$uriprefix&path=$(urlencode "$path")&content-length=$filelength&content-md5=$filemd5&slice-md5=$fileslicemd5&content-crc32=$filecrc32&ondup=$ondup"

;;

add_task)

uri="$uriprefix&save_path=$(urlencode "$path")&source_url=$(urlencode "$link")"

;;

query_task)

optype=

uri="$uriprefix&task_ids=$taskid&op_type=$optype"

;;

list_task)

start=

limit=

asc= #

task_info= #

uri="$uriprefix&start=$start&limit=$limit&asc=$asc&need_task_info=$task_info"

;;

cancel_task)

uri="$uriprefix&task_id=$taskid"

;;

listrecycle)

uri="$uriprefix&start=0"

;;

restore)

uri="$uriprefix&fs_id=$fsid"

;;

esac

######################################

## Building arguments for CURL ##

curlopts=(

-X$req

"${CURL_DEFAULT_ARGS[@]}"

)

case $method in

upload)

curlopts=("${curlopts[@]}" -F file=@"$filepath" "$uri")

;;

download)

curlopts=("${curlopts[@]}" -o "$savefile" "$uri")

;;

delete|add_task|list_task|query_task|cancel_task|restore|mkdir|rapidupload)

curlopts=(

    "${curlopts[@]}"

    -H "Content-Length: 0"

    -Ls "$uri"

)

;;

move|copy)

curlopts=(

    "${curlopts[@]}"

    -H "Content-Type: application/json"

    --data param="{\"list\":[{\"from\":\"$from\",\"to\":\"$to\"}]}"

    -Ls "$uri"

)

;;

*)

curlopts=("${curlopts[@]}" -Ls "$uri")

;;

esac

#################################

### Start the operation ###

# echo "${curlopts[@]}"; exit

case $method in

download)

curl "${curlopts[@]}"

;;

*)

curl "${curlopts[@]}"|

tokenize|pprint|

sed '/^ *"path": /s/^\( *"path": "\)'"$(regexescape "$pathprefix")"'\(.*"\)$/\1\2/'|

compact -ftbm | autounit

;;

esac

##########################

使用，假设保存成文件名bdcurl的执行文件：
如果想不用浏览器，而是直接在命令行授权给bypy来使用的话，需要在浏览器里登录 过百毒云，然后在cookie里找到BDUSS（可以用firebug，或者使用sqlite打开浏览器的数据库也能找到），把内容复制到脚本开头的 BDUSS变量里（这个有效期比较长，所以基本只复制一次就可以了，以后token过期后重新认证时可以省略这步），
bdcurl auth
等一小会就会输出token id，把这个段字符串复制到脚本开头的TOKEN变量里，然后就可以使用了。

如果不方便获取BDUSS，就需要把终端输出的网址粘贴到浏览器里获取一个code，然后把code替换
https://bypy-tianze.rhcloud.com/auth?code=$code&redirect_uri=oob&
里面$code的部分，获得最后的access_token放到脚本开始的TOKEN变量里。
然后命令行运行：
bdcurl h 看帮助

bdcurl u file1 down #上传文件到down文件夹
bdcurl f file1 down #快速上传文件（只向服务器查询校验信息，如果百度云上有相同的文件，则直接在建立该文件，不需上传，如果返回error_code，则百毒云上没有相同文件，需要使用bdcurl u上传文件）
bdcurl d down/file1 # 下载指定路径的文件
bdcurl del down/file1 # 删除指定路径的文件
bdcurl l down # 列出 down目录下的所有文件

暂时不支持批量上传下载，毕竟百毒本身没有这样的API，所谓批量得在客户端实现，但是bash来实现相对会麻烦很多，因为没有很好的解析json的方法，如果使用python，perl之类的辅助解析，那就不如干脆直接用python或perl写了。

另外输出结果使用一个compact函数对其做一定程度的简化，其简化程度可通过参数指定，脚本里默认(compact -ftbm, 倒数第四行)使用最大程度简化（如，去除花括号和方括号，去除fs_id，md5校验码，size文件大小，mtime和ctime创建/修改时间 等项目，如果传递参数里去掉-s则会保留size项），可以自行参看脚本里定义compact函数接收参数的说明，选择你合适的输出简化程度