在MacOS 10.13.3的Anaconda环境中部署QGIS的可行性咨询

First off, let's be clear: this is a moderately challenging task—not something a total beginner can knock out in an afternoon, but totally doable if you have some experience with compiling software and managing Python environments. Here's a breakdown of the key hurdles and what to expect:

1. Dependency Library Alignment #

QGIS relies on a ton of third-party libraries: geospatial core ones like GDAL, PROJ, and GEOS, plus UI tools like Qt and PyQt. The catch is these are often built to work with system Python or python.org's distribution, not Anaconda. To make this work:

• You'll need to ensure all dependencies are either installed via conda or compiled specifically against your Anaconda environment. Mismatched library paths (like linking to system GDAL instead of conda's) will throw runtime errors.
• Some less common dependencies might not exist in conda's repositories, forcing you to compile them from source—adding more steps to the process.
• You’ll have to prioritize Anaconda’s environment variables during compilation to avoid accidentally pulling in system-level tools or libraries.

2. Tweaking Build Configurations #

QGIS uses CMake for its build pipeline, so you’ll need to adjust several CMake flags to point it to your Anaconda Python:

• Specify PYTHON_EXECUTABLE to your Anaconda environment’s Python binary (e.g., ~/opt/anaconda3/envs/your_qgis_env/bin/python).
• Set PYTHON_INCLUDE_DIR and PYTHON_LIBRARY to the corresponding paths in your Anaconda installation.
• You might also need to manually set CMAKE_PREFIX_PATH to help CMake find conda-installed dependencies like GDAL or Qt.

3. Version Compatibility Headaches #

Your macOS 10.13.3 (High Sierra) is pretty old, and newer QGIS versions often require macOS 10.14 or later. This means:

• You’ll need to track down a QGIS source version that supports macOS 10.13 (likely an early 3.x release like 3.8 or 3.10).
• Make sure your Anaconda Python version matches what that QGIS release expects—for example, QGIS 3.10 requires Python 3.6+, so your conda env should meet that.

4. Debugging and Troubleshooting Time #

Compiling QGIS will almost certainly throw errors along the way:

• Missing dynamic libraries, version mismatches, or broken Python module imports.
• Qt/PyQt UI compilation failures if versions don’t line up.
  Each issue will require digging through build logs, adjusting environment variables, or tweaking CMake settings. If you’re new to this, expect to spend a lot of time troubleshooting each roadblock.

Ways to Simplify the Process #

If you want to avoid a full-from-scratch compile, try these shortcuts:

• First, install all core dependencies via conda: conda install gdal proj geos qt pyqt—this eliminates the need to compile most of the heavy lifting yourself.
• Follow QGIS’s official macOS build guide, but replace all references to system/python.org Python with your Anaconda paths.
• Look for existing community scripts or CMake configs that target Anaconda on macOS—you can adapt these instead of starting from zero.

Overall, this task requires patience and some familiarity with build tools and environment management, but it’s definitely achievable if you’re willing to work through each step.

内容的提问来源于stack exchange，提问作者user1256923