summary |
shortlog |
changelog |
graph |
tags |
bookmarks |
branches |
files |
changeset |
raw | gz |
help

author | nipkow |

Thu, 11 Aug 2022 13:23:00 +0200 | |

changeset 75805 | 3581dcee70db |

parent 75804 | dd04e81172a8 |

child 75806 | 2b106aae897c |

child 75807 | b0394e7d43ea |

removing the [simp] attribute breaks too many AFP entries severely

--- a/src/HOL/Library/NList.thy Thu Aug 11 11:57:19 2022 +0200 +++ b/src/HOL/Library/NList.thy Thu Aug 11 13:23:00 2022 +0200 @@ -14,20 +14,23 @@ lemma nlistsI: "\<lbrakk> size xs = n; set xs \<subseteq> A \<rbrakk> \<Longrightarrow> xs \<in> nlists n A" by (simp add: nlists_def) -lemma nlistsE_length (*rm simp del*): "xs \<in> nlists n A \<Longrightarrow> size xs = n" +text \<open>These [simp] attributes are double-edged. + Many proofs in Jinja rely on it but they can degrade performance.\<close> + +lemma nlistsE_length [simp]: "xs \<in> nlists n A \<Longrightarrow> size xs = n" by (simp add: nlists_def) lemma less_lengthI: "\<lbrakk> xs \<in> nlists n A; p < n \<rbrakk> \<Longrightarrow> p < size xs" -by (simp add: nlistsE_length) +by (simp) -lemma nlistsE_set[simp](*rm simp del*): "xs \<in> nlists n A \<Longrightarrow> set xs \<subseteq> A" +lemma nlistsE_set[simp]: "xs \<in> nlists n A \<Longrightarrow> set xs \<subseteq> A" unfolding nlists_def by (simp) lemma nlists_mono: assumes "A \<subseteq> B" shows "nlists n A \<subseteq> nlists n B" proof fix xs assume "xs \<in> nlists n A" - then obtain size: "size xs = n" and inA: "set xs \<subseteq> A" by (simp add:nlistsE_length) + then obtain size: "size xs = n" and inA: "set xs \<subseteq> A" by (simp) with assms have "set xs \<subseteq> B" by simp with size show "xs \<in> nlists n B" by(clarsimp intro!: nlistsI) qed