Mirror of forum.nim-lang.org

3581 :: SIGSEGV: Illegal storage access. (Attempt to read from nil?) 3

• nimboolean (orginal) [2018-02-25T05:27:37+01:00] view original

  The code

  import htmlparser
  import xmltree
  import strtabs
  import strutils
  
  
  let allhtml:File = open("allhtml.txt")
  var line: string
  
  while allhtml.readline(line):
    echo "HTML FILE: ", line
    var html = loadHTML(line.strip())
    for a in html.findall("a"):
      var url = a.attrs.getOrDefault("href", nil)
      if not url.isNil and url.startswith("/"):
        echo "$1\t$2" % [line, url]
  
  close(allhtml)
  
  Run

  The error:

  
  Traceback (most recent call last)
  rewrite.nim(26)          rewrite
  strtabs.nim(133)         getOrDefault
  strtabs.nim(104)         rawGet
  strtabs.nim(85)          myhash
  SIGSEGV: Illegal storage access. (Attempt to read from nil?)
  Run

---

• StasB (orginal) [2018-02-25T05:37:29+01:00] view original

  It looks like the second argument to getOrDefault is supposed to be a non-nil string.

---

• nimboolean (orginal) [2018-02-25T05:41:06+01:00] view original

  nope. it did not make any difference.

  import htmlparser
  import xmltree
  import strtabs
  import strutils
  
  
  let allhtml:File = open("allhtml.txt")
  var line: string
  var count = 0
  while allhtml.readline(line):
    echo "HTML file: ", count, line
    var html = loadHTML(line.strip())
    for a in html.findall("a"):
      var url = a.attrs.getOrDefault("href", "/")
      if not url.isNil and url.startswith("/"):
        echo "$1\t$2" % [line, url]
    count += 1
  
  close(allhtml)
  Run

---

• nimboolean (orginal) [2018-02-25T05:57:32+01:00] view original

  ISSUE RESOLVED.

  I fixed the error by checking for attrsLen property like this:

  import htmlparser
  import xmltree
  import strtabs
  import strutils
  
  
  let allhtml:File = open("allhtml.txt")
  var line: string
  var count = 0
  while allhtml.readline(line):
    echo "HTML file: ", count, line
    let html = loadHTML(line.strip())
    for a in html.findall("a"):
      if a.attrsLen > 0:
        let url = a.attrs.getOrDefault("href")
        if not url.isNil:
          if url.startswith("/"):
            echo "$1\t$2" % [line, url]
    count += 1
  
  close(allhtml)
  Run

---