Enterprise: add support for expanded context window to Gemini 1.5 models

[abeatrix]

CONTEXT: https://sourcegraph.slack.com/archives/C05AGQYD528/p1718323087549179?thread_ts=1718316482.993009&cid=C05AGQYD528

CLOSE https://linear.app/sourcegraph/issue/CODY-2356/add-gemini-to-the-list

• This change adds support for additional chat models with expanded context windows, including 'claude-3-opus', 'claude-3-sonnet', 'gpt-4o', 'gpt-4-turbo', and 'gemini-1.5'.
• The isModelWithExtendedContextWindowSupport function has been updated to check for these model substrings, and the getEnterpriseOutputLimit function has been simplified to use the new function for determining the appropriate output limit.

Test plan

N/A.

Adding a new string to the current list. No feature was changed.

[aramaraju]

@abeatrix This approach won't work for GPT-4o through Azure OpenAI. Here's what the config looks like docs:

[abeatrix]

@abeatrix This approach won't work for GPT-4o through Azure OpenAI. Here's what the config looks like docs:

This PR is for adding support to Gemini 1.5 models for Cody Gateway Enterprise users though.

[abeatrix]

@abeatrix This approach won't work for GPT-4o through Azure OpenAI. Here's what the config looks like docs:

The model ID for GPT-4o through Azure OpenAI is also listed as gpt-4o in their doc, so shouldn't this work?