HTMLValidator : Java Glossary

CSE (Computer Science and Engineering) HTMLValidator Pro 17.02 Last revised/verified: 2017年03月22日is the HTML (Hypertext Markup Language) validator I use to check my *.html web pages for syntax errors. It also checks *.css style sheets. It is produced by A I Internet Solutions aka htmlvalidator.com. Albert Wiersh of A.I. has been extremely helpful over the years whenever I had a problem or suggestions for new features. CSE does not currently stand for or mean anything (other than just being part of the name), but when the program was originally named, it stood for CSE

Why Validate?

HTMLValidator is extremely configurable. You can get it to ignore picky errors for now, then when you have the big problems with your site cleaned up, turn them back on. You can also make it super picky, so that your HTML is easier to process with programs, or to gradually migrate your site to XHTML (extensible Hypertext Markup Language).

Trial/Buy

HTMLValidator comes in three versions. See this comparison chart to help you decide which version you need.

Configuring

Here’s how I configure it to demand balancing </li> tags.

1. Click Options ⇒ Configuration Editor ⇒ Edit ⇒ Tag Names.
2. On the top right, Select li (020) with the drop down box on the right (not the box on the left) as the Tag Name Being Configured
3. On the right, select the Required end tag radio button.
4. You’ll probably want to repeat this for all li tags whose context is distinguished with a number.
5. Save the configuration file to a different filename to make sure it doesn’t get overwritten if the CSE HTML Validator is reinstalled or updated.

In Options ⇒ Validation Engine Option ⇒ Options

• I set check for high ASCII (American Standard Code for Information Interchange) character on.
• I set require lower case tags and attributes. Consistency never hurts.
• I set required quoted attribute values. This consistency helps later with any search/replace or one-off text processing programs.

Set the max number of errors and warnings before aborting at Tools ⇒ Tools Options ⇒ Message Output.

The configuring options are squirreled in over seven different places. Check out:

• Tools ⇒ Tools Options: mostly checkboxes
• Tools ⇒ Tools Options ⇒ Config file ⇒ Edit configuration: which tags are mandatory.
• Tools ⇒ Batch Wizard ⇒ options ⇒ Batch Wizard options: mostly checkboxes. Max files in a batch.
• Options ⇒ Editor Options; mostly how the editor looks.
• Options ⇒ Validator Engine Options ⇒ options: same as Tools ⇒ Tools Options
• Options ⇒ Validator Engine Options ⇒ edit configuration files: e.g. configure HTMLTidy.
• Options ⇒ Configuration Editor ⇒ edit configuration: same as Tools ⇒ Tools Options ⇒ Config file ⇒ Edit configuration

Unfortunately, when you install a new version of the program, usually your old customisations are lost unless you make special provisions.

The program itself installs as: X:\Program Files (x86)\HTMLValidator170\cse170.exe

Other files in:
"C:\Users\user\AppData\Roaming\AI Internet Solutions\CSE HTML Validator\17.0\", Files you can configure include:

• batchreporttemplate.css
• cssinsertercfg.txt
• htmlvalV170.cfg: I suggest backing up this file in some permanent data directory so it will be backed up and won’t get lost on reinstall.
• javascriptinsertercfg.txt
• msgidcfg.xml
• options.veo: I suggest backing up this file in some permanent directory so it will be backed up and won’t get lost on reinstall.
• phpinsertercfg.txt
• taginsertercfg.txt
• tidyprofiles.xml
• userinsertercfg.txt

& and &amp;

Gotchas

• Make sure you uncheck the batch wizard option After validating a target list, set process to ‘No’ for the targets with no errors, warnings or bad links. Otherwise your validations will gradually stop working.
• The check for high ASCII characters is done without regard to tags. Unfortunately, this means you can’t use <!--cseignore--> to suppress high ASCII error messages.
• Uncheck Options ⇒ Validator Engine Options ⇒ Validator Engine Options ⇒ Enable JavaScript Lint messages to disable JavaScript checking. The problem is if your HTML contains VBScript, the validator gets confused and thinks it is JavaScript and complains wherever the syntax differs from JavaScript.

Command Line Options

I have not found a way to load a file into the editor and validate it. I wish there were ways to load a file, validate and exit if all is well. I also wish that were true for a batch, but when there were some files in the batch with problems, they should be loaded in the editor, automatically, ready to correct.

Tips

• To configure which browser to use for the Batch Reports, configure the browser in Windows settings.
• Validation does not replace proof-reading. There are many sorts of errors that validation cannot catch including missing or scrambled text, using the wrong word, leaving off the & on the front of entities, ugly layouts, text with the wrong css class, a spelling error inadvertently added to the list of good words, inconsistencies in style… Nothing beats just perusing the pages from time to time in a variety of browsers looking for anomalies.
• Use copy/paste the file name in the batch validation browser report to revalidate the bad document in regular mode, where you can correct the errors in its editor.
• You can download spelling dictionaries for additional languages as *.adm files. 11 languages come built-in. Just unzip them and copy them into X:\Program Files (x86)\HTMLValidator170\Dictionary\.
• Invoke validator on a single file like this: [フレーム]
• If there are a large number of documents needing correction, copy/paste the list from the batch validation browser report to a file and tidy them to look like this:

  -o E:\mindprod\jgloss\google.html
  -o E:\mindprod\jgloss\htmlcheat.html
  -o E:\mindprod\jgloss\intellij.html
  -o E:\mindprod\jgloss\jetpurchase.html

  and save them in a file badlist.txt. Then submit the entire file with:

  rem submit list of -o (open) commands to HTMLValidator command line processor
  %HTMLValidatorDir\cmdlineprocessor.exe -f badlist.txt
  Rem -30-

• Use the recent feature the batch validator to only check files that have changed within n days where n can be a fraction. This will save needlessly checking files over and over. I now run a global check of all recently changed files several times a day. This helps me keep on top of any errors. Create a batch descriptor to validate your whole site, but limit it to changes within the last day. Run this before every upload. It will complete quite quickly and this habit will prevent you from ever uploading bad HTML.
• Invoke the batch processor from the command processor like this: [フレーム]

Validate your Webpages Free

Custom Features

I have talked the HTMLValidator people into implementing some new features, but they are not in the standard edition. You have to ask for them specially.

• This script track of which documents it has already validated and when. Then it could transparently avoid revalidating documents that have not changed since the last validation. You could ask to revalidate everything and it would be clever enough just to revalidate what has changed since the last validation, even if different documents were validated in different batches. It would look as if it had revalidated everything very quickly. You handle this by customising the following javascript and telling the batch validator to invoke it by putting the name of the script file in Target List Options ⇒ user functions file. [フレーム]
• Documents that fail to pass a batch validates are automatically loaded into the editor for correction. Without the feature, you have to type their names one by one to load them for correction. You handle this telling the batch validator to invoke the following script file by putting the name of the script file in Target List Options ⇒ user functions file. [フレーム]

RFE (Request For Enhancement)

Here is my wish list of features primarily I would like to make HTMLValidator better at spell checking. I have asked many times before and been rejected, however, I am eternally optimistic these low priority wishes will eventually percolate to the top. The most important items are at the top of the list.

1. Treat entities embedded inside words as if there were the equivalent Unicode before looking them up in the spell check dictionary.
2. Allow some way to mark blocks of text with magic comments or styles to request they not be spell checked.
3. Pay attention to the lang markup and use the correct language dictionary. If there is no corresponding dictionary, suppress checking.
4. Batch check a set of documents for spelling. Auto close those without error. Leave those with errors open in the editor.
5. Spell checker than can handle tags embedded inside words, e.g. when acronyms are explained, e.g. URL
6. Command line operations fail if you move to a different directory without shutting HTMLValidator down first. It should let you leave it running all the time while you move from directory to directory feeding it files to validate.
7. Consolidate all the configuration menus in one place and reorganise them so finding something to configure is easier to find. Assign configuration options short names rather than numbers.
8. When doing a spell check run, it would be nice if HTMLValidator allowed you to you to type in a word not in the dictionary to correct. Currently you must stop the spell checker, correct the problem with editor, then restart the spell checker.

Scripts

HTMLValidator allows you to write little chunks of JavaScript and insert them to customise the way HTMLValidator works. I use two scripts. The first keep track of which of my files have already been successfully validated and have not since changed. Why a do a batch validation, such files don:t have to be revalidated. It makes the whole process at least an order of magnitude faster.

This script automatically loads files into the editor if they fail validation.

They call it Tag Name Programming Language You can read up on how it works with the batch wizard.