[SOLVED] parsing words from a specific line

Issue

Following is python implementation to extract specific fields from the following input file. For the line starting with MEMRANGE, I would like to extract the values corresponding to keys BASEVALUE, INSTANCE and SLAVEBUSINTERFACE and print the result in the following format. value[INSTANCE].value[SLAVEBUSINTERFACE] \tab =value[BASEVALUE]

Similarly, for the line starting with PORT, I would like to extract the values corresponding to keys CLKFREQUENCY and SIGNAME and print the result as value[SIGNAME] \tab =value[CLKFREQUENCY]

input file

       <MEMRANGE ADDRESSBLOCK="HP0_DDR_LOW" BASENAME="C_BASEADDR" BASEVALUE="0x00000000" HIGHNAME="C_HIGHADDR" HIGHVALUE="0x3FFFFFFF" INSTANCE="trimap_0" IS_DATA="TRUE" IS_INSTRUCTION="TRUE" MASTERBUSINTERFACE="M_AXI" MEMTYPE="MEMORY" SLAVEBUSINTERFACE="S_AXI_HP0_hmo"/>

        <PORT CLKFREQUENCY="90005550" DIR="I" NAME="ACLK" SIGIS="clk" SIGNAME="SYS_clk_wiz_0_clk_out1">

        <MEMRANGE ADDRESSBLOCK="HP0_DDR_LOW" BASENAME="C_BASEADDR" BASEVALUE="0x00000000" HIGHNAME="C_HIGHADDR" HIGHVALUE="0x3FFFFFFF" INSTANCE="trimap_0" IS_DATA="TRUE" IS_INSTRUCTION="TRUE" MASTERBUSINTERFACE="M_AXI_DRAM0" MEMTYPE="MEMORY" SLAVEBUSINTERFACE="S_AXI_HP0_hmo"/>
        <PORT CLKFREQUENCY="90005550" DIR="I" NAME="ACLK" SIGIS="clk" SIGNAME="SYS_clk_wiz_0_clk_out1">

I have tested the individual functions and they work fine. When I try to write the output of the functions to the output file, only the first function is executed and the second function doesn't execute.

import re

def populate_ip_address(input_file, output_file):
    print(" > writing ip address")
    for line in input_file:
        match_address = re.match(r'.*MEMRANGE .*BASEVALUE="(\w+)\".* INSTANCE="(\w+)\".* SLAVEBUSINTERFACE="(\w+)\"',line)

        if match_address:
            newline1= "\n%s.%s\t\t\t\t\t= %s" % (match_address.group(2), match_address.group(3), match_address.group(1))
            output_file.write(newline1)

def populate_clock_frequency(input_file, output_file):
    print(" > writing clock_frequency")
    for line in input_file:
        match_clock = re.match(r'.*PORT .*CLKFREQUENCY="(\w+)\".* SIGNAME="(\w+)\"',line)

        if match_clock:
            print(" >> writing clock_frequency")
            newline2= "\n%s \t= %s" % (match_clock.group(2), match_clock.group(1))
            output_file.write(newline2)

input_file = open("test.txt", "r")
with open('new.txt','w') as output_file:
    populate_clock_frequency(input_file, output_file)
    populate_ip_address(input_file, output_file)

expected

SYS_clk_wiz_0_clk_out1  = 90005550
SYS_clk_wiz_0_clk_out1  = 90005550
trimap_0.S_AXI_HP0_hmo                  = 0x00000000
trimap_0.S_AXI_HP0_hmo                  = 0x00000000

current output

SYS_clk_wiz_0_clk_out1  = 90005550
SYS_clk_wiz_0_clk_out1  = 90005550

Am I doing something wrong in the file access? I am also open to answers doing the above parsing in sed, awk or grep.

Solution

Assumptions:

• lines of interest (1st field is <MEMRANGE or <PORT ) reside on a single line (ie, they do not span multiple lines)

General approach:

• split (space delimited) fields on equal sign (=)
• if 1st sub-field is one of our 5 desired attributes then ...
• split the 2nd sub-field on double quotes (") and ...
• store the attribute/value pair in an array
• print the contents of the array

One awk idea:

awk '
BEGIN               { n=split("INSTANCE|SLAVEBUSINTERFACE|BASEVALUE|SIGNAME|CLKFREQUENCY",a,"|")

                      for (i=1;i<=n;i++)                  # build associative array where
                          attrib_list[a[i]]               # indices are the desired attributes
                    }

/<MEMRANGE |<PORT / { delete val
                      for (i=2;i<=NF;i++) {               # loop through space-delimited fields
                          split($i,a,"=")                 # split <attrib>="<value>" on equal sign
                          if (a[1] in attrib_list) {      # if this is one of the desired attributes then ...
                             split(a[2],b,"\"")           # split "<value>" on double quote and ...
                             val[a[1]]=b[2]               # save: val[<attrib>]=<value>
                          }
                      }
                      if ($1=="<MEMRANGE")
                         printf "%s.%s\t= %s\n", val["INSTANCE"], val["SLAVEBUSINTERFACE"], val["BASEVALUE"]
                      else
                         printf "%s\t= %s\n", val["SIGNAME"], val["CLKFREQUENCY"]
                     }
' input_file

This generates:

trimap_0.S_AXI_HP0_hmo  = 0x00000000
SYS_clk_wiz_0_clk_out1  = 90005550
trimap_0.S_AXI_HP0_hmo  = 0x00000000
SYS_clk_wiz_0_clk_out1  = 90005550

If order is important OP can pipe the results to sort, eg:

$ awk '... awk script ...' input_file | sort
SYS_clk_wiz_0_clk_out1  = 90005550
SYS_clk_wiz_0_clk_out1  = 90005550
trimap_0.S_AXI_HP0_hmo  = 0x00000000
trimap_0.S_AXI_HP0_hmo  = 0x00000000

NOTES:

• output is based on the single \t delimiter as mentioned in the textual description
• OP's expected output appears to show 2 extra tabs (\t) for the trimap* lines while OP's current code appears to show 5x tabs (\t)
• OP can modify the printf format to generate the desired number of tabs (\t)

Answered By - markp-fuso

Answer Checked By - Mary Flores (WPSolving Volunteer)