代码之家 › 专栏 › 技术社区 › rockyroad

从日志文件中提取单词

sed awk bash

0

rockyroad · 技术社区 · 8 年前

我试图从日志文件中提取作业id,但在bash中提取它们时遇到问题。我试过用sed。

我的日志文件如下:

> 2018-06-16 02:39:39,331 INFO  org.apache.flink.client.cli.CliFrontend 
> - Running 'list' command.
> 2018-06-16 02:39:39,641 INFO  org.apache.flink.runtime.rest.RestClient                      
> - Rest client endpoint started.
> 2018-06-16 02:39:39,741 INFO  org.apache.flink.client.cli.CliFrontend                       
> - Waiting for response...
>  Waiting for response...
> 2018-06-16 02:39:39,953 INFO  org.apache.flink.client.cli.CliFrontend                       
> - Successfully retrieved list of jobs
> ------------------ Running/Restarting Jobs -------------------
> 15.06.2018 18:49:44 : 1280dfd7b1de4c74cacf9515f371844b : jETTY HTTP Server -> servlet with content decompress -> pull from
> collections -> CSV to Avro encode -> Kafka publish (RUNNING)
> 16.06.2018 02:37:07 : aa7a691fa6c3f1ad619b6c0c4425ba1e : jETTY HTTP Server -> servlet with content decompress -> pull from
> collections -> CSV to Avro encode ->  Kafka publish (RUNNING)
> --------------------------------------------------------------
> 2018-06-16 02:39:39,956 INFO  org.apache.flink.runtime.rest.RestClient                      
> - Shutting down rest endpoint.
> 2018-06-16 02:39:39,957 INFO  org.apache.flink.runtime.rest.RestClient                      
> - Rest endpoint shutdown complete.

我正在使用以下代码提取包含jobid的行:

extractRestResponse=`cat logFile.txt`
echo "extractRestResponse: "$extractRestResponse

w1="------------------ Running/Restarting Jobs -------------------"
w2="--------------------------------------------------------------"
extractRunningJobs="sed -e 's/.*'"$w1"'\(.*\)'"$w2"'.*/\1/' <<< $extractRestResponse"
runningJobs=`eval $extractRunningJobs`
echo "running jobs :"$runningJobs

但是这没有给我任何结果。另外,我注意到在打印extractrestresponse变量时,所有新行都会丢失。

我也试过使用这个命令,但没有给出任何结果:

extractRestResponse="sed -n '/"$w1"/,/"$w2"/{//!p}' logFile.txt"

3 回复 | 直到 8 年前

1

SLePort 8 年前

用SED:

sed -n '/^-* Running\/Restarting Jobs -*/,/^--*/{//!p;}' logFile.txt

说明:

默认情况下,在应用命令后,输入行将回显到标准输出。这个 -n 标志禁止此行为
/^-* Running\/Restarting Jobs -*/,/^--*/ :匹配起始行 ^-* Running\/Restarting Jobs -* 高达 ^--* (包括在内)
//!p; :打印与地址匹配的行以外的行

2

1

karakfa 8 年前

awk 救命啊!

awk '/^-+$/{f=0} f; /^-+ Running\/Restarting Jobs -+$/{f=1}' logfile

3

0

builder-7000 8 年前

您可以改进原始替换:

sed -e 's/.*'"$w1"'\(.*\)'"$w2"'.*/\1/' <<< $extractRestResponse

通过使用 @ 作为分隔符:

sed -n "s@.*$w1\(.*\)$w2.*@\1@p" <<< $extractRestResponse

输出是介于 $w1 和 $w2 :

> 15.06.2018 18:49:44 : 1280dfd7b1de4c74cacf9515f371844b : jETTY HTTP Server -> servlet with content decompress -> pull from > collections -> CSV to Avro encode -> Kafka publish (RUNNING) > 16.06.2018 02:37:07 : aa7a691fa6c3f1ad619b6c0c4425ba1e : jETTY HTTP Server -> servlet with content decompress -> pull from > collections -> CSV to Avro encode -> Kafka publish (RUNNING) >