In an older wiki post (which I know is outdated...but still...) located here: https://wiki.apache.org/solr/FileBasedSpellChecker, the end of the first paragraph (the intro) indicates that "...it isn't all that hard to create an index from a file and have it weight the terms."
I've been thinking about this, and it would be PERFECT for what we need (because we HAVE a dictionary from a different search engine WITH weights on it that have been crafted/tweaked over a long period of time). However, as I've read the spellcheck documentation, I'm really not certain what this person (Mark Bennett) was talking about. Question #1: Has anyone pursued this possibility? If so, can you offer some insight as to how you approached this? Question #2: I'd be grateful for someone's "hint" as to where to start looking to try to create this workaround if you have some experience with it. I'm guessing that the only way to do this is to dive into the File based spell check dictionary code and then clone/modify what is there. I was just wondering if I was missing something or if someone was aware of a quicker way to scale this technological hill. I realize I could dissect the format of the spellcheck dictionary and load it as I desire...but I'm a bit concerned about this as this is for a long-lived project that is expected to live a lot longer, and this approach seems to be a bit "brittle." It will break the first time Apache makes the slightest change to the structure of the spell check index. If anyone has any insights to offer I'd appreciate it. Thanks. Peter S. Lee
