I have a file called index.html and I grep '<td class="headerCovTableEntryLo">' index.html getting this:
<td class="headerCovTableEntryLo">39.2 %</td>
<td class="headerCovTableEntryLo">56.6 %</td>
I need to get only 39.2. How can I achieve this ?
You can parse the HTML using an XML parser such as xmlstarlet.
xmlstarlet fo -H page.html |
xmlstarlet sel -t -v '//td[@class="headerCovTableEntryLo"][1]' -n 2>/dev/null |
tr -dc '[:digit:].\n'
Output
39.2
The first invocation of xmlstarlet parses the HTML and converts it as best it can to XML. The second invocation parses that XML and extracts the first <td/> element value with an attribute class matching headerCovTableEntryLo. The final tr removes the space and percent characters from the string (actually, it removes everything except digits, the dot, and newline).
If you need a more precise match you will need either to make the //... path more specific or post more of the HTML.
If you want to avoid the call out to tr you can return the element value upto but excluding the first space (so 39.2 % will be returned as 39.2):
xmlstarlet fo -H page.html |
xmlstarlet sel -t -v 'substring-before(//td[@class="headerCovTableEntryLo"][1], " ")' -n 2>/dev/null
If you have GNU grep , you can use the -P option to enable Perl Compatible Regular Expresions which give us the \K regex command which means "ignore anything matched up to this point". Combined with -m1 to stop at the first match and -o to print only the matching portion of the line, you can do:
$ grep -m1 -oP '<td class="headerCovTableEntryLo">\K[0-9.]' file
39.2
Or, similarly:
$ grep -m1 -oP '(?<=<td class="headerCovTableEntryLo">)[0-9.]+' file
39.2
Alternatively, with awk:
$ awk -F'[> ]' '/<td class="headerCovTableEntryLo">/{print $3; exit}' file
39.2
[0-9.] which I was already using in the second one, but had forgotten to fix the first one.
With GNU grep, you can use -m 1 to exit after the first match. e.g.
grep -m 1 '<td class="headerCovTableEntryLo">' index.html
From man grep:
-m NUM, --max-count=NUM
Stop reading a file after
NUMmatching lines. If the input is standard input from a regular file, andNUMmatching lines are output, grep ensures that the standard input is positioned to just after the last matching line before exiting, regardless of the presence of trailing context lines. This enables a calling process to resume a search.When grep stops after NUM matching lines, it outputs any trailing context lines.
When the
-c or--countoption is also used, grep does not output a count greater than NUM. When the-vor--invert-match` option is also used, grep stops after outputting NUM non-matching lines.
To then extract just the value, pipe it into sed. e.g.
$ grep -m 1 '<td class="headerCovTableEntryLo">' index.html |
sed -e 's/^[^>]*>//; s/ %.*//'
39.2
Or, forget about grep and do the whole thing in sed:
$ sed -ne '/<td class="headerCovTableEntryLo">/ {s/^[^>]*>//; s/ %.*//p;q}' index.html
39.2
You really ought to use a HTML parser, though. Using regular expressions alone on structured data like HTML, XML, JSON, etc is doomed to failure.
grep above into sed -e 's/^[^>]*>//; s/ %.*//
index.html is a gcov & lcov & genhtml output. I am sure this won't break for a long time.
pup for html parsing, instead of core utils. pup 'td[class="headerCovTableEntryLo"] -f index.html | head -1 ought to fix it.
To complete all the tools grep, awk and sed:
sed -En '/<td class="headerCovTableEntryLo">([0-9.]+).*/{s//\1/p;q}'
<td class="headerCovTableEntryLo">39.2 %</td>or just39.2?grep?