例子：搜索用户表 user 中字段 phone 包含 520 的行。
SQL
和 pymysql 等库一样的用：
keyword = '520'
conn.execute('select * from user where phone like "%%%s%%";' % keyword)
conn.execute('select * from user where phone like "%%%s%%";', keyword)
.like 方法
q = session.query(model.User.id, model.User.phone).filter(model.User.phone.like(f'%{keyword}%'))
qs = qs.all()
# print(qs.statement)
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE "user".phone LIKE :phone_1
对应的大小写不敏感方法有 ilike (lower("user".phone) LIKE lower(:phone_1))
还有：not_like, not_ilike
.contains 方法
print(session.query(model.User.id, model.User.phone).filter(model.User.phone.contains(keyword)).statement)
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE ("user".phone LIKE '%' || :phone_1 || '%')
.regexp_match 方法（1.4 新增）
对应的是 MySQL 支持的 REGEXP 操作符。
print(session.query(model.User.id, model.User.phone).filter(model.User.phone.regexp_match(keyword)).statement)
session.query(model.User.id, model.User.phone).filter(model.User.phone.regexp_match(keyword)).all()
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE "user".phone <regexp> :phone_1
.startswith 和 .endswith
print(session.query(model.User.id, model.User.phone).filter(model.User.phone.startswith(keyword)).statement)
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE ("user".phone LIKE :phone_1 || '%')
print(session.query(model.User.id, model.User.phone).filter(model.User.phone.endswith(keyword)).statement)
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE ("user".phone LIKE '%' || :phone_1)
.match 方法
对应的是数据库的 MATCH (col1,col2,...) AGAINST (expr [search_modifier]) 全文索引方法。
对单字段同样可用，不过需要先建立 FULLTEXT 索引。
print(session.query(model.User.id, model.User.phone).filter(model.User.phone.match(keyword)).statement)
session.query(model.User.id, model.User.phone).filter(model.User.phone.match(keyword)).all()
# SELECT "user".id, "user".phone
# FROM "user"
# WHERE "user".phone MATCH :phone_1
