The Lazy Programmer's XML Parser

Question

Background

You are working as a programmer for a car sales company. Your task for this week is to program an XML parser that takes in data about available models from different car manufacturers, and pretty-prints information about the newest models. Luckily for you, the testing department has provided only one test case! The faster you can write code that passes it, the more time you have for procrastination during the rest of the week.

Input

Your input is exactly this piece of XML data, supplied by the testing department. It contains data about some car manufacturers, their series of cars, and the models in these series. You may assume a trailing newline.

<?xml version="1.0" ?>
<products>
  <manufacturer name="Test Manufacturer 1">
    <series title="Supercar" code="S1">
      <model>
        <name>Road Czar</name>
        <code>C</code>
        <year>2011</year>
      </model>
      <model>
        <name>Ubervehicle</name>
        <code>U</code>
        <year>2013</year>
      </model>
      <model>
        <name>Incredibulus</name>
        <code>I</code>
        <year>2015</year>
      </model>
      <model>
        <name>Model 1</name>
        <code>01</code>
        <year>2010</year>
      </model>
    </series>
    <series title="Test series 22" code="Test">
      <model>
        <name>Test model asdafds</name>
        <code>TT</code>
        <year>2014</year>
      </model>
    </series>
  </manufacturer>
  <manufacturer name="Car Corporation">
    <series title="Corporation Car" code="CC">
      <model>
        <name>First and Only Model</name>
        <code>FOM</code>
        <year>2012</year>
      </model>
    </series>
  </manufacturer>
  <manufacturer name="Second Test Manufacturer">
    <series title="AAAAAAAAAAAAAA" code="D">
      <model>
        <name>Some older model</name>
        <code>O</code>
        <year>2011</year>
      </model>
      <model>
        <name>The newest model</name>
        <code>N</code>
        <year>2014</year>
      </model>
    </series>
    <series title="BBBBBBBBBBBBBBB" code="asdf">
      <model>
        <name>Another newest model here</name>
        <code>TT</code>
        <year>2015</year>
      </model>
    </series>
  </manufacturer>
</products>

Output

Your output is this string. It lists the car manufacturers in alphabetical order, followed by a colon and the number of series they make. Under each manufacturer, it lists the series name, model name, and code of each of their models, starting from the newest and going backward by year. Trailing whitespace and line breaks are acceptable, as long as your output looks similar to this when printed.

Car Corporation: 1 series
  Corporation Car, First and Only Model (CC-FOM)
Second Test Manufacturer: 2 series
  BBBBBBBBBBBBBBB, Another newest model here (asdf-TT)
  AAAAAAAAAAAAAA, The newest model (D-N)
  AAAAAAAAAAAAAA, Some older model (D-O)
Test Manufacturer 1: 2 series
  Supercar, Incredibulus (S1-I)
  Test series 22, Test model asdafds (Test-TT)
  Supercar, Ubervehicle (S1-U)
  Supercar, Road Czar (S1-C)
  Supercar, Model 1 (S1-01)

Rules and Scoring

You may write either a function or full program. The lowest byte count wins, and standard loopholes are disallowed.

Note that the input is fixed: you don't need to support any other inputs than the one given here. Your program is allowed to return nonsense or even crash if the input is modified in any way. You may also ignore the input and hard-code the output, if desired. However, you may not use XML or HTML parser libraries or built-ins.

Answer 1

CJam, 109 107 bytes

"rzn ¸À¨ 4T\$D"S/q"<>"'"er'"/
f{\:i2/_E5bf.+.+\ff=)1<_s,9/+":  series
"2/.+\"  ,  ()
-"2/Z4e\f.\}

Note that four of the characters in the string at the beginning are unprintable.

Try it online in the CJam interpreter.

Idea

This is basically a hardcode that splits the input at all occurrences of <, > and ", selects specific chunks and interleaves them with the remaining parts of the output.

After splitting the input, the chunks at indexes 110, 114 and 122 are Car Corporation, Corporation Car and First and Only Model. The codes for series and name can be found at indexes 116 and 126, which can be calculated by adding 2 and 4 to the indexes of the names. Finally, the number of series is the length of the string Car Corporation divided by 9 (obviously).

Thus, we encode the part of the output that corresponds to this manufacturer as [114 122 110] or rather the string "rzn".

Code

e# Split the string at spaces.

"rzn ¸À¨ 4T\$D"S/

e# After replacing characters with their code points, this will be the result:
e# [[114 122 110] [184 192 144 168 144 152 140] [12 52 84 92 12 36 12 20 12 68 8]]

e# Read from STDIN, replace < and > with " and split at double quotes.

q"<>"'"er'"/ 

f{       e# For each chunk of the split string, push the split input; then:
\:i2/    e# Replace characters with character codes and split into chunks of
         e# length 2.
_E5b     e# Copy the result and push [2 4].
f.+      e# Replace each pair [I J] in the copy with [I+2 J+4].
.+       e# Vectorized concatenation.
         e# For the first chunk, we've achieved
         e# [114 122 110] -> [[114 122] [110]]
         e#               -> [[114 122] [110]] [[116 126] [110 112 4]]
         e#               -> [[114 122 116 126] [110 112 4]]
\ff=     e# Replace each integer with the chunk of the split input at that index.
)1<      e# Pop the first chunk and reduce it to the first string.
_s,9/    e# Divide that strings length by 9 and append the integer to the array.
":  series
"2/      e# Push [": " " s" "er" "ie" "s\n"].
.+       e# Vectorized concatenation; pushes the manufacturer line.
\        e# Swap the remaining strings on top of the stack.
"  ,  ()
-"2/     e# Push ["  " ", " " (" ")\n" "-"].
Z4e\     e# Swap the chunks at indexes 3 and 4.
\f.\     e# Interleave both arrays of strings.
}

Answer 2

Bubblegum, 227 225 bytes

0000000: 6d 50 45 b6 02 31 10 dc cf 29 6a 87 eb 92 1d ee 0e 07  mPE..1...)j.......
0000012: 08 93 1e 3c e1 45 be 9d fe 37 ae bd 2b 7d 95 54 85 41  ...<.E...7..+}.T.A
0000024: 55 9b 83 36 c2 ad b5 2a a1 00 4b 66 4d 36 c0 23 0f f6  U..6...*..KfM6.#..
0000036: a5 d1 58 1b eb 20 94 c4 50 ed 7e d1 d7 92 76 88 57 ab  ..X.. ..P.~...v.W.
0000048: 99 c6 b0 9f 08 a6 14 6a 96 66 c4 9e be 50 3e 12 a1 f3  .......j.f...P>...
000005a: 86 4c 09 c5 7b 67 e5 f9 d2 28 2b ed 56 64 a0 e8 9b 83  .L..{g...(+.Vd....
000006c: d8 9f 3a 99 20 c4 85 95 51 66 36 4b 70 ac fc 74 69 cc  ..:. ...Qf6Kp..ti.
000007e: 56 f4 9c 88 d7 32 83 4f c6 a9 de 13 f4 4e 92 b9 1b 87  V....2.O.....N....
0000090: 89 e0 6d 24 0a 4f 33 a7 fe 40 26 e4 37 a3 ad 42 43 72  ..m$.O3..@&.7..BCr
00000a2: bd f0 3b 6f 11 9f 16 32 ed 04 eb a7 fc d9 8d 62 91 f7  ..;o...2.......b..
00000b4: dc 97 f0 6a 11 49 f6 1e b9 cb fc 7b dd 7c 41 e6 8b 56  ...j.I.....{.|A..V
00000c6: eb 70 47 a7 b6 f9 b3 3c d1 42 a2 fa 27 cc 49 ac 3e 89  .pG....<.B..'.I.>.
00000d8: 97 ff 2e 9c a4 7c 21 f1 0f                             .....|!..

This isn't very competitive, but I just couldn't resist to post my first Bubblegum answer to an actual kolmogorov-complexity challenge.

The hexdump can be reversed with xxd -r -c 18 > xml.bg.

The code completely ignores the input. Compression has been done with zopfli, which uses the DEFLATE format but obtains a better ratio than (g)zip.

Thanks to @Sp3000 for -2 bytes!

Answer 3

sed, 449 bytes

Assumes sed to be run with the -nr options.

/\?|<p/d;:M
/<ma/{s/.*"(.*)".*/Q\1: X series/;/C/ s/X/1/;s/X/2/;H;d}
/<se/{s/.*"([^"]*)".*"([^"]*)".*/@\1!\2/;H;d}
/<mo/{
G;s/.*@(.*)*$/\1/;x;s/@(.*)*$//;x;:A N
/<\/m/!bA
s/\n/!/g;s/  +//g;s|<[/a-z]*>||g;s/(.*)!(.*)!(.*)!(.*)!(.*)!/%\1, \3 (\2-\4)@\1!\2/;H}
/<\/se/{x;s/\n*@.*$//;x}
$!{n;bM}
x;s/\n//g;s/Q(.*)Q(.*)%(.*)Q(.*)/\2\n  \3\n\4\n\1/
s/%(.*)%(.*)%(.*)\n(.*)%(.*)%(.*)%(.*)%(.*)%(.*)/\n  \3\n  \2\n  \1\n\4\n  \7\n  \9\n  \6\n  \5\n  \8/;p

Ungolfed version:

# Remove first 2 lines
/\?|<p/ d

:M
#manufacturer
/<ma/ {
    s/.*"(.*)".*/Q\1: X series/
    #Car corp
    /C/ s/X/1/
    #other
    s/X/2/
    H
    d
}

#series: store in hold buffer. (treating the hold buffer as a list, with @ as a delimeter)
/<se/{
    s/.*"([^"]*)".*"([^"]*)".*/@\1!\2/
    H
    d
}
/<mo/ {
    # pull series from hold buffer
    G
    s/.*@(.*)*$/\1/

    # remove series from hold buffer
    x
    s/@(.*)*$//
    x

    # Concatenate model into one line
    :A N
    /<\/m/ !{
        bA
    }
    s/\n/!/g

    # Remove junk
    s/  +//g
    s|<[/a-z]*>||g

    # Append formatted line to hold buffer, replace series at the end
    s/(.*)!(.*)!(.*)!(.*)!(.*)!/%\1, \3 (\2-\4)@\1!\2/
    H
}
/<\/se/ {
    #pop series name
    x
    s/\n*@.*$//
    x
}

$ ! {
    n
    b M
}
# end of loop

x
s/\n//g

# "sort"
s/Q(.*)Q(.*)%(.*)Q(.*)/\2\n  \3\n\4\n\1/
s/%(.*)%(.*)%(.*)\n(.*)%(.*)%(.*)%(.*)%(.*)%(.*)/\n  \3\n  \2\n  \1\n\4\n  \7\n  \9\n  \6\n  \5\n  \8/
p

Answer 4

Bash, 388 368 365

base64 -d<<<H4sICKVS9FUCA2hlcmUAbVFBasMwELwH8oc92mBD5GNuqUogB9dQOw9QpDUWKFJYWS3t6yvJtI3T7k07szOzKy4IuKObIzFrZ/fAwCNp9NsN3APABVVw1ORnEFZBZ80HtE6hgYLz+ti15XbTo3QRGzCSWmHDKOQcCGkPzZ3q07oqOFg3T0hg8T1NXrNqbCAUwquxHoYyzR1WVcEw4XqkeK5f/mX27orgjIoeP8wuMv8EBbaO2ocbkkybn6wkVPoSTPBQ9Kw+ZaessPChaarlvXjE6GJUkZx63zv8Cp4vSG84aWkw650f8FcnFPDP+D0J5Q/ocnmWsR0rvwC2OTuexgEAAA==|zcat

Small test because:

$ bash ./go_procrastination.sh cars.xml
Car Corporation: 1 series
  Corporation Car, First and Only Model (CC-FOM)
Second Test Manufacturer: 2 series
  BBBBBBBBBBBBBBB, Another newest model here (asdf-TT)
  AAAAAAAAAAAAAA, The newest model (D-N)
  AAAAAAAAAAAAAA, Some older model (D-O)
Test Manufacturer 1: 2 series
  Supercar, Incredibulus (S1-I)
  Test series 22, Test model asdafds (Test-TT)
  Supercar, Ubervehicle (S1-U)
  Supercar, Road Czar (S1-C)
  Supercar, Model 1 (S1-01)