Cannot start Heritrix with Windows 10

[machawk1]

Related to #155 but for Windows.

I compiled WAIL on Windows 10 from fc82057 then installed the latest Java from Oracle manually. When selecting the FIX heritrix button, the console reports that it will be opened in a new Windows.

The new window displays this:

Java version 1.8.0_201
Java SE Runtime Environment (build 1.8.0_201-b09)
Java HotSpot Client VM (build 25.201-b09, mixed mode)

[machawk1]

Also relevant, fc82057 uses Heritrix 3.2.0 and per #345, a new version if available but is not a drop-in replacement per discussion in #345. We may want to replicate the commands to Heritrix that WAIL issues with an isolated version of Heritrix run on Windows 10 to verify compatibility between Heritrix and a newer version of Java than expected on macOS.

[machawk1]

See https://webarchive.jira.com/browse/HER-2085, per internetarchive/heritrix3#129 this might be fixed by updating Heritrix as in #345.

[machawk1]

To Get It Working™, let's bundle the Java 1.7 JDK like we have on macOS. If I recall, this was one barrier in Heritrix -- the Java security APIs that are used in Heritrix were removed in Java 1.8.

[machawk1]

With Java 7 SDK bundled for Windows, some of the Java checks are unnecessary. Unfortunately, if Java 8 is also installed on the system, it seems to get precedence over the bundled Java. Programatically adding the environment variables, as we do in ensureEnvironmentVariablesAreSet(), still allows Heritrix to use the system Java (version 8 as of a fresh Windows 10).

[machawk1]

This may have to do with how Heritrix is invoked on startup. Check Heritrix.launchHeritrix().

Running the invocation: C:\wail\bundledApps\heritrix-3.2.0\\bin\heritrix.cmd -a lorem:ipsum from the command line (understandable) still causes the system Java to be used outside of the context of WAIL. I believe that the environment variables like JAVA_HOME can be used by preceding this invocation, which might be a solution to remedy WAIL using the system Java.

EDIT

heritrix.cmd also provides this option within the startup script:

::  Optional environment variables
:: 
::  JAVA_HOME        Point at a JDK install to use.

EDIT2

...or perhaps not. There does not appear to be a flag to send this value to Heritrix but instead, it ought to be EXPORTed prior to calling heritrix.cmd.

[machawk1]

Having the WAIL-Win code in WAILConfig.py also set jreHome and javeHome prevents the Java8 from being used.

...but then the command to heritrix.cmd states "The system cannot find the path specified.\nPress any key to continue . . ." without giving more information on the command.

[machawk1]

Heritrix now runs in Windows in 90cadc1 \o/

...but hitting the API to create and build scripts still seems problematic to the inaccessibility as reported on the CLI.

<urlopen error [WinError 10049] The requested address is not valid in its context>

This may be related to #435 with regard to Windows not being configured with localhost access to 0.0.0.0 by default.

As suspected, when Heritrix is running on Windows, the web interface is accessible using https://localhost:8443/ and not https://0.0.0.0:8443/ by default.

[machawk1]

Wayback appears to use the system Java by default as well per Tomcat startup.

[machawk1]

Modified the Windows catalina startup script in bd8c30d